1: % submitted to PRB on 21/4/01

2: %\documentstyle[12pt,epsfig]{article}

3: \documentclass[12pt]{article}

4: \usepackage{epsfig,amssymb}

5: 

6: \setlength{\topmargin}{0.2cm}

7: \raggedbottom

8: \abovedisplayskip=3mm

9: \belowdisplayskip=3mm

10: \abovedisplayshortskip=0mm

11: \belowdisplayshortskip=2mm

12: \setlength{\baselineskip}{16pt}

13: \setlength{\evensidemargin}{0pt}

14: \setlength{\oddsidemargin}{0in}

15: \setlength{\parskip}{0.13cm}

16: \setlength{\textwidth}{16truecm}

17: \setlength{\textheight}{20cm}

18: \baselineskip=20pt

19: 

20: \newcommand\beq{\begin{equation}}

21: \newcommand\eeq{\end{equation}}

22: \newcommand\bea{\begin{eqnarray}}

23: \newcommand\eea{\end{eqnarray}}

24: \newcommand{\nonum}{\nonumber}

25: \newcommand\sqpi{\sqrt{\pi}}

26: \newcommand\px{\partial_x}

27: \newcommand\pt{\partial_\tau}

28: \newcommand\bo{{\bar \omega}}

29: \newcommand\on{{\bar \omega}_n}

30: \newcommand\tphi{\tilde\phi}

31: \newcommand\tchi{\tilde\chi}

32: \newcommand\ua{\uparrow}

33: \newcommand\da{\downarrow}

34: \newcommand\phiou{\phi_{1\uparrow}}

35: \newcommand\phiod{\phi_{1\downarrow}}

36: \newcommand\phitu{\phi_{2\uparrow}}

37: \newcommand\phitd{\phi_{2\downarrow}}

38: \newcommand\tphiou{{\tilde \phi}_{1\uparrow}}

39: \newcommand\tphiod{{\tilde \phi}_{1\downarrow}}

40: \newcommand\tphitu{{\tilde \phi}_{2\uparrow}}

41: \newcommand\tphitd{{\tilde \phi}_{2\downarrow}}

42: \newcommand\ttheta{\tilde\theta}

43: \newcommand\tk{\tilde k}

44: 

45: \begin{document}

46: 

47: \centerline{\Large Conductance through contact barriers of a finite length 

48: quantum wire}

49: 

50: \vskip .5 true cm

51: \centerline{\bf Siddhartha Lal$^1$, Sumathi Rao$^2$ and Diptiman Sen$^1$} 

52: \vskip .5 true cm

53: 

54: \centerline{\it $^1$ Centre for Theoretical Studies} 

55: \centerline{\it Indian Institute of Science, Bangalore 560012, India} 

56: \vskip .5 true cm

57: 

58: \centerline{\it $^2$ Harish-Chandra Research Institute}

59: \centerline{\it Chhatnag Road, Jhusi, Allahabad 211019, India}

60: \vskip .5 true cm

61: 

62: \begin{abstract}

63: 

64: We use the technique of bosonization to understand a variety of recent 

65: experimental results on the conductivity of a quantum wire. The quantum wire 

66: is taken to be a finite-length Luttinger liquid connected on two sides to 

67: semi-infinite Fermi liquids through contacts. The contacts are modeled as 

68: (short) Luttinger liquids bounded by localized one-body potentials. We use 

69: effective actions and the renormalization group to study the effects of

70: electronic interactions within the wire, the length of the wire, 

71: finite temperature and a magnetic field on the conductivity. We explain the 

72: deviations of the conductivity away from $2Ne^2 /h$ in wires which are not

73: too short as arising from renormalization effects caused by the repulsive 

74: interactions. We also explain the universal conductance corrections observed

75: in different channels at higher temperatures. We study the effects of an 

76: external magnetic field on electronic transport through this system and 

77: explain why odd and even spin split bands show different renormalizations 

78: from the universal conductance values. We discuss the case of resonant 

79: transmission and of the possibility of producing a spin-valve which only

80: allows electrons of one value of the spin to go through. We compare our

81: results for the conductance corrections with experimental observations.

82: We also propose an experimental test of our model of the contact regions.

83: 

84: \end{abstract}

85: \vskip .5 true cm

86: 

87: ~PACS number: 72.10.-d, 85.30.Vw, 71.10.Pm, 73.40.Cg

88: 

89: \newpage

90: 

91: \section{\bf Introduction}

92: 

93: With the rapid advances made in the fabrication of high mobility

94: semiconductor heterojunctions, these systems have provided the 

95: setting for the discovery of several new phenomena in quantum 

96: systems. Popular examples include mesoscopic systems like 

97: quantum dots, quantum wires, and the two-dimensional electron gas samples 

98: in which the quantum Hall effects are observed \cite{datta}. 

99: In particular, quantum wires are created by the electrostatic gating of 

100: two-dimensional electron gases (2DEG) (with typical densities of 

101: $n_{2DEG} \sim 0.5 - 6 \times 10^{15} m^{-2}$) in the inversion layer of GaAs 

102: heterostructures. These GaAs samples typically have a very high mobility 

103: (typically, $\mu \sim 3 - 8 \times 10^{2} m^{2} V^{-1} s^{-1}$) because there 

104: is very little disorder in them; the mean free path of an electron in the 

105: 2DEG is of the order of $\lambda_{MF} \sim 5 - 20\mu m$. 

106: This makes it possible to create ballistic channels a few microns in length 

107: for studying electron transport in such wires, especially 

108: at low temperatures when the 

109: thermal de Broglie wavelength of the electron is comparable to the channel 

110: length. Furthermore, since it is possible to maintain a low carrier 

111: concentration in these wires, it becomes possible for transport to take place 

112: through only a few channels or even a single channel.

113: 

114: Thus, several observations [2-12]

115: of the quantization of the conductance in electron 

116: transport through such channels have been made over the last two decades. More 

117: recently, several new ways have been found to produce such channels and this 

118: has led to even more precise experimental studies. This has brought into focus 

119: novel aspects of electron transport in such channels, not all of which 

120: are understood as yet. 

121: 

122: Let us first briefly review some of the recent

123: experimental findings in quantum wires.

124: The first striking observation is that of a number of flat plateaus in the dc 

125: conductance which are separated by steps of roughly the same value 

126: \cite{yacoby}

127: \beq

128: g = N{\tilde g}~\frac{2e^2}{h} ~,

129: \eeq

130: where $N$ denotes positive integers starting from one. The factor 

131: ${\tilde g} < 1$ and is found to vary with the length of the quantum 

132: wire and the temperature; it has been seen to be as low as $0.75$. In fact, 

133: the plateaus tend towards $N(2e^2/h)$ as either the temperature is raised or 

134: the length of the quantum wire is shortened \cite{tarucha,yacoby}. This seems 

135: to imply a uniform renormalization of the plateau heights for each channel 

136: in the quantum wire as a function of wire length and temperature 

137: \cite{yacoby,facer,liang2,reilly}. Also, 

138: the flatness of the plateaus appears to indicate an insensitivity to the 

139: electron density in the channel. Furthermore, kinks are observed on the 

140: rise of the conductivity to some of the lowest plateaus. One such kink 

141: has been named the ``$0.7$ effect" \cite{thomas,liang,reilly}. These kinks are 

142: seen to wash away 

143: quickly with increasing temperature \cite{thomas,liang} and an external 

144: magnetic field placed in-plane and parallel to the channels \cite{liang}. Also, 

145: upon increasing such a magnetic field, a splitting of the conductance steps 

146: is observed together with an {\it odd-even} effect of the renormalization of 

147: the plateau heights, with the odd and even plateau heights being 

148: renormalized by smaller and larger amounts respectively. Finally, at very 

149: high magnetic field, another kink is seen to arise near the first 

150: spin-split plateau. 

151: 

152: Several of these experimental observations have found no satisfactory 

153: explanation till date. It is the purpose of this work to provide a consistent 

154: framework within which most of these observations of transport in quantum 

155: wires can be explained. We will rely upon several concepts and techniques that 

156: have become popular in the study of interacting mesoscopic systems. These 

157: include the concept of bosonization, effective actions and the renormalization 

158: group (RG) \cite{haldane,kane,furusaki1}. To be precise, we will employ 

159: these techniques and ideas to understand the low energy transport properties 

160: of ballistic electrons in a finite quantum wire attached to two semi-infinite 

161: Fermi leads \cite{safi,maslov,furusaki}, but with a difference; the contacts 

162: of the quantum wire with its leads will themselves be modeled as short 

163: quantum wires with junction barriers at either end. The properties of the 

164: contacts are unaffected by the 

165: external gate voltage which causes the formation of the discrete sub-bands in 

166: the quantum wire. The junction barriers will be 

167: modeled as localized $\delta$-functions to account for the back-scattering of 

168: electrons due to the imperfect coupling between the quantum wire and the 2DEG 

169: reservoirs; these barriers will renormalize the conductance as 

170: observed in the experiments. We will also study the effect of external electric 

171: and magnetic fields on this system. The properties of such a model for the 

172: quantum wire will be seen to account for several of the experimental 

173: observations mentioned above, as well as predict the possibility of some more 

174: interesting observations in future experiments in these systems. It should 

175: be stated here that a possible mechanism for some of the experimental 

176: observations \cite{yacoby} has been proposed in 

177: Ref. \cite{alekseev}; this is based on the 

178: anomalously enhanced back-scattering of electrons entering the 2DEG reservoir 

179: from the quantum wire due to the formation of Friedel oscillations of the 

180: electron density near the edges of the reservoirs, and it neglects 

181: interactions between the electrons in the quantum wire. Our model, however, 

182: attempts to understand these observations keeping in mind the importance 

183: of electron-electron interactions, barrier back-scattering, finite

184: temperature and magnetic field as well as all

185: length scales in the quantum wire system. 

186: 

187: The paper is organized as follows. In Sec. 2, we discuss the basics of the 

188: model outlined above. We show that the model can be described by a

189: $K_L$-$K_C$-$K_W$ Luttinger model

190: \cite{sumathi} with three different interaction parameters 

191: in the lead ($K_L$), the contacts ($K_C$) and the wire ($K_W$). By assuming

192: that the electron-electron interactions get screened out rapidly as one goes 

193: from one region to another, we get in a natural way the existence of 

194: localized barriers at the junctions. We then discuss how our model goes

195: beyond the concept of {\it ideal} contact resistances as studied in the 

196: Landauer-Buttiker formalism. In Sec. 3, we study the effective action of 

197: our model in the presence of external electric fields after integrating 

198: out all bosonic fields except those at the boundaries between the various 

199: regions - the leads, the contacts and the wire. Depending on the

200: relative sizes of the contacts and the wire, we define two regimes

201: - a) the quantum wire (QW) limit, where

202: the length of the wire is much greater than the contact, and b) the 

203: quantum point contact (QPC) limit, where the length of the contacts is 

204: much greater than the length of the wire. We then study the symmetries

205: of the effective action to determine when resonant transmission is possible 

206: as a function of a tunable gate voltage. All of this is first done

207: for spinless fermions and subsequently, we give the modifications of

208: the results for spinful fermions. In Sec. 4, we explicitly compute

209: the corrections to the conductance due to the barriers at finite

210: temperature ($T$) and for a finite length of the wire ($l$) and for finite 

211: contact lengths ($d$). We compute the frequency dependent Green's

212: functions of the model, in the different frequency regimes and use the Kubo 

213: formula to compute the conductance.

214: We also show how these results could have been anticipated from the 

215: renormalization group (RG) equations for the barrier strengths. In Sec. 5, 

216: we study our model in the presence of an external in-plane

217: magnetic field. Using RG 

218: methods, we show that the spin-up and spin-down electrons see different 

219: barrier heights at the junctions, and use this idea to explain the 

220: odd-even effect mentioned earlier. We also outline all the 

221: possible resonances that can be seen under such conditions. We point out

222: the possibility of producing a spin-valve at moderate magnetic fields.

223: In addition, we compute the conductance of our model 

224: and discuss its qualitative features as a function of the strength of

225: the magnetic field. In Sec. 6, we compare the features of 

226: the conductance expressions obtained with the observations made in 

227: various experiments for transport in quantum wires with and without an external 

228: magnetic field. We find that our model is applicable to a large class of 

229: experiments and gives a unified and qualitatively correct explanation of all 

230: of them. In particular, our model gives a possible explanation 

231: for the uniform renormalization of all the conductance steps seen in several

232: experiments. We also explain the odd-even effect seen in

233: experiments in the presence of magnetic field. In addition, we propose

234: more precise experimental tests of our model. 

235: Finally, we end in Sec. 7, with a summary of all the new results in our paper,

236: and outline further investigations that are possible. 

237: 

238: \section{\bf The Model}

239: 

240: In this section, we will study the Tomonaga-Luttinger liquid (TLL)

241: model \cite{haldane} of a quantum wire of finite length with no disorder, 

242: which is connected to the two 2DEG reservoirs modeled as two semi-infinite 

243: Fermi leads through two contact regions. The contacts are modeled

244: as short quantum wires with the junctions at either end modeled as 

245: $\delta$-function barriers. The inter-electron interactions in the

246: system, and hence the parameter $K$ which characterizes the interactions, vary 

247: abruptly at each of the junctions. Hence, we study a

248: $K_L$-$K_C$-$K_W$-$K_C$-$K_W$ model (see Fig. 1). The motivation for the 

249: above model is as follows. The electrons in the 2DEG are basically 

250: free, and hence, in the equivalent

251: 1D model, they are modeled as semi-infinite leads with 

252: Luttinger parameter $K_L=1$. This can be understood as follows: if each end 

253: of the quantum wire is approximated by a point, only those 

254: electrons in the 2DEG which are in a zero angular momentum state (with respect 

255: to the appropriate end) can enter (or leave) the wire. Thus, the wave function 

256: of such a state has the radial coordinate as its only variable and we may, 

257: therefore, model the 2DEG as noninteracting 1D systems lying 

258: on either side of the quantum wire. The electron velocity in the leads 

259: $v_L$ is given by the Fermi velocity of the 2DEG electrons in the reservoirs 

260: $v_F = \sqrt{2E_{F2D}/m}$. On the other hand, the externally applied 

261: gate voltage $V_G$ is applied over a small region and this leads to the 

262: formation of several discrete sub-bands where the electrons feel the 

263: transverse confinement potential produced by $V_G$. This region is the 

264: one-dimensional quantum wire where the density of electrons is controlled by 

265: the gate voltage. The lowest energies $E_s$ in each sub-band are given by the 

266: discrete energy levels for the transverse confinement potential (and can 

267: therefore be shifted by changing $V_G$) \cite{buttiker}. The Fermi energy in 

268: the $s^{\rm th}$ channel is given by $E_{F1D} = E_{F2D} - E_s$. A channel is 

269: open when $E_{F1D} > 0$; the electron velocity $v_{W}(e)$ in the channel 

270: is then related to the 2DEG Fermi velocity $v_F$ by $v_W (e) = {\sqrt {v_F^2 - 

271: 2 E_s /m}}$. In this gate voltage constricted region, the electrons will be 

272: considered as interacting via a short range (Coulomb-like) repulsion. 

273: Thus each discrete channel is modeled by a separate TLL. Let us, for the 

274: moment, consider one such channel with an interaction parameter $K_W$ and 

275: quasiparticle velocity $v_W$. 

276: 

277: The contacts represent the regions where the geometry changes from 

278: two-dimensional (2D) to 1D. In these regions of 

279: changing geometry, interactions between 

280: the electrons are likely to be very important; thus we model

281: the contact region as a Luttinger liquid with $K=K_C$. However, the gate 

282: voltage $V_G$ is unlikely to affect the properties of the electrons in these 

283: regions as the discrete sub-bands form a little deeper inside the wire. 

284: We choose different parameters $K_W$ for the wire and $K_C$ for the contact, 

285: because it is not obvious that inter-electrons interactions 

286: within the quantum wire will be the same as in the contact. The density of 

287: electrons in the quantum wire is controlled by the gate voltage, whereas the 

288: density of electrons in the contacts is controlled by the density of the 2DEG 

289: at or near the Fermi energy. Hence, we expect $K_C$ to be independent of 

290: $V_G$, but $K_W$ is dependent on $E_{F1D}$, which, in turn depends on $V_G$.

291: We will also show below that the change in the 

292: inter-electron interactions in each of the lead, contact and quantum wire 

293: regions gives rise to barrier-like back-scattering of the electrons. 

294: 

295: Simpler versions of this model (but without junction barriers and without 

296: contacts) have been studied by several authors \cite{safi,maslov,furusaki} 

297: who found perfect conductance through the TLL channel which is independent 

298: of the inter-electron interactions. Perfect conductance is also seen in 

299: several of the experiments \cite{wees,tarucha,yacoby,thomas,liang}. 

300: In the opposite limit, the model of a finite quantum wire connected to the 

301: two reservoirs by tunneling through very large barriers has also been 

302: studied \cite{fabrizio}. The idea of modeling 2DEG reservoirs by 1D 

303: noninteracting Fermi leads has also earlier been 

304: employed in studies of the fractional quantum Hall effect edge states coupled 

305: to Fermi liquids through a tunneling term in the Hamiltonian \cite{chamon}. 

306: Some studies of disordered quantum wires in such a model (again with perfect 

307: junctions) have also been conducted and the corrections to the conductance 

308: because of back-scattering impurities found \cite{safilong,maslovlong}. 

309: The continuity of 

310: the results found in these studies (which have quantum wires of a finite 

311: length) with those found earlier for infinite quantum wires \cite{kane} has 

312: also been established \cite{furusaki,safilong}. 

313: 

314: The main difference between our model and the earlier studies of the

315: quantum wire is that here we explicitly model the contacts as short

316: TLL wires bounded by junction barriers on either end and whose properties 

317: are unaffected by the gate voltage $V_G$. As we will discuss later, an 

318: experiment performed recently \cite{picciotto}

319: has conclusively shown the existence of a 

320: region (of an appreciable length of $2-6\mu m$) in between the quantum 

321: wire and the 2DEG reservoirs which leads to the back-scattering of 2DEG 

322: electrons entering the quantum wire. Furthermore, the idea that the 

323: properties of a one-dimensional system are determined by the Fermi 

324: energy of the 2DEG reservoirs has been used in Ref. \cite{matveev} to 

325: study the quantum point contact. In addition, we assume that the changes 

326: in the inter-electron interactions take place abruptly in going from the 

327: contacts into the quantum wire and that all inter-electron interactions get 

328: screened out very quickly in going from the contacts into the leads. It can, 

329: however, be shown that a smoother variation of the interaction parameter $K$ 

330: upon going from the quantum wire into the contacts and in going from the 

331: contacts into the noninteracting leads does not affect any of the transport 

332: properties in the $\omega \to 0$ (dc)

333: limit as long as we have no barriers of any kind in the 

334: system. We will now show that changes in the inter-electron 

335: interactions at the lead-contact and contact-quantum wire junctions give rise 

336: to barrier-like terms in the Hamiltonian of the system; the existence of these

337: terms is mentioned briefly in the work of Safi and Schulz \cite{safilong}. 

338: This is, however, only one reason why the junctions between the 

339: $1$D channel and its leads can cause the back-scattering of electrons; another 

340: reason is clearly the change in geometry 

341: in going from the 2DEG reservoirs into the 1D channel. 

342: This cause for the drop in the conductance of the channel has earlier been 

343: studied within the purview of the Landauer-Buttiker formalism; see

344: \cite{yacimry} and references therein. 

345: 

346: Let us begin by studying the simpler case of a quantum wire (in which electrons 

347: are interacting with each other) connected directly to the noninteracting, 

348: semi-infinite leads without any intermediate contact regions. Then there is 

349: only a single change in inter-electron interactions from zero in the leads to 

350: a finite value in the quantum wire. The kinetic part of the Hamiltonian for 

351: this system of interacting spinless electrons in the quantum wire when 

352: expressed in terms of the bosonic field $\phi(x)$ and its canonically 

353: conjugate momentum $\Pi(x)= \partial_t \phi/v_F$ is given by

354: \beq 

355: H_0 = \frac{1}{2} \int dx v_F [ \Pi(x)^2 + (\partial_x\phi(x))^2 ] ~,

356: \eeq

357: where $v_F$ is the Fermi velocity of the electrons in the channel. The part of 

358: the Hamiltonian which characterizes the short-ranged density-density 

359: interactions between the electrons in a 1D channel of length $l$ is given by

360: \beq

361: H_{int} = \int_0^l dx \int_0^l dy ~{\cal U}(x,y) \rho(x) \rho(y) ~,

362: \eeq

363: where ${\cal U}(x,y)$ characterizes the strength of the 

364: density-density interactions between the electrons, and $\rho(x)$ is the 

365: electronic density at the point $x$. Using a truncated form of the Haldane 

366: representation for the electronic density in terms of the bosonic field 

367: $\phi(x)$ \cite{haldane}, the density $\rho(x)$ is given by 

368: \beq

369: \rho(x) = \frac{1}{\sqpi}\partial_x\bar\phi(x) ~[c_0 + 

370: 2c_1\cos(2\sqpi \bar\phi(x))]~,

371: \eeq

372: where $\bar \phi(x) = \phi(x) + k_Fx/\sqpi$, $c_0 = 1$, $c_1 = 

373: \Lambda /(2k_F)$, and $k_F$ is the Fermi wave vector. $\Lambda$ is the 

374: ultraviolet cutoff ($\Lambda < O(E_{F1D})$); 

375: it is the energy limit up to which the linearization of

376: the bands and hence bosonization is expected to be applicable.

377: If we now characterize the short range inter-electron interactions 

378: by ${\cal U}(x,y) = {\cal U}_0 \delta(x-y)$, 

379: then we can substitute the expressions for the density and the inter-electron 

380: interaction into the interaction term in the Hamiltonian. This gives us

381: \bea

382: H_{int} &=& {\cal U}_0 \int_0^l dx ~[\{\partial_x(\frac{c_0}{\sqpi}

383: \bar\phi)\}^2 ~-~2\partial_x(\frac{c_0}{\sqpi}\bar\phi)\partial_x (

384: \frac{c_1}{\sqpi} \sin (2\sqpi \bar\phi)) \nonum \\

385: & & ~~~~~~~~~~~~~~+~ \{\partial_x (\frac{c_1}{\sqpi}

386: \sin (2\sqpi\bar\phi) )\}^2] \nonum \\

387: &=& {\cal U}_0 \int_0^l dx ~[\frac{c_0^2}{\pi}(\partial_x\bar\phi)^2

388: ~+~2\frac{c_0^2k_F}{\pi}\partial_x\phi~+~2\frac{c_0c_1k_F}{\sqpi}

389: \partial_x\sin (2\sqpi \bar\phi) \nonum \\

390: & & ~~~~~~~~~~~~+~2\frac{c_0c_1}{\sqpi}\partial_x\phi\partial_x

391: \sin (2\sqpi \bar\phi) ~+~2\frac{c_1^2}{\pi}(\partial_x\bar\phi)^2

392: (\cos(4\sqpi\bar\phi) +1)] \nonum \\

393: &=& {\cal U}_0 \int_0^l dx ~[(\frac{c_0^2+2c_1^2}{\pi})

394: (\partial_x\phi)^2

395: ~+~2k_F(\frac{c_0^2+2c_1^2}{\pi})\partial_x\phi~+~2\frac{c_0c_1k_F}{\sqpi}

396: \partial_x\sin (2\sqpi\bar\phi) \nonum \\

397: & & ~~~~~~~~~~~~+~2\frac{c_0c_1}{\sqpi}\partial_x\phi\partial_x

398: \sin (2\sqpi\bar\phi) ~+~2\frac{c_1^2}{\pi}(\partial_x\bar\phi)^2

399: \cos(4\sqpi\bar\phi) ].

400: \eea

401: We can now simplify this expression by noting that several of the terms above 

402: contain rapidly oscillating factors of $\cos(k_Fx)$ or $\sin(k_Fx)$ which 

403: make those terms vanish upon performing the 

404: integration (unless we are at very specific fillings of the electron density).

405: Thus, we can ignore the fifth term straightaway. The first term can 

406: be added to a similar term in $H_0$ where it renormalizes the velocity and 

407: introduces an interaction parameter $K$. 

408: The second term is a chemical potential term and that too can be accounted for 

409: by shifting the field $\phi$ accordingly. The third term is clearly a boundary 

410: term, and it gives us two barrier like terms at $x=0$ and $x=l$, with

411: \bea

412: H_{barrier} &=& 2{\cal U}_0\frac{c_0c_1k_F}{\sqpi} \int_0^l dx ~

413: \partial_x \sin (2\sqpi\bar\phi) \nonum \\

414: &=& 2{\cal U}_0\frac{c_0c_1k_F}{\sqpi}~

415: [\sin (2\sqpi\phi(l)+ 2k_Fl)~-~\sin (2\sqpi\phi(0)) ].

416: \eea

417: Finally, the fourth term can also be rewritten as

418: \bea

419: H_{int,4} &=& 4{\cal U}_0\frac{c_0c_1}{\sqpi} \int_0^l dx ~(\partial_x

420: \phi)^2 \cos (2\sqpi\bar\phi)~+~2{\cal U}_0\frac{c_0c_1k_F}{\sqpi} \int_0^l dx 

421: ~\partial_x\sin (2\sqpi\bar\phi) \nonum \\

422: & & -~4{\cal U}_0\frac{c_0c_1k_F^2}{\sqpi} \int_0^l dx ~\cos (2\sqpi\bar\phi)~,

423: \eea

424: in which the first and third terms again vanish because they contain rapidly 

425: oscillating factors within the integrals, and the second term adds on to 

426: $H_{barrier}$ exactly. All this finally gives us two $\delta$-function 

427: barriers at the junctions of the quantum wire with its Fermi liquid leads as 

428: \beq

429: H_{barrier} = 4{\cal U}_0 \frac{c_0c_1k_F}{\sqpi}~[ \sin(2\sqpi\phi(l)+ 

430: 2k_Fl)~-~\sin (2\sqpi\phi(0))] ~.

431: \eeq

432: The extension of the derivation given above to our model with two 

433: intermediate contact regions where the inter-electron interactions are 

434: ${\cal U}(x,y) = {\cal U}_1 \delta(x-y)$ 

435: (i.e., different from that in the quantum wire) is straightforward, 

436: and it yields four barrier terms: two at the junctions of the 

437: contacts with the leads, and two at the junctions of the contacts with the 

438: quantum wire. It is also very likely that the inner two barriers are much 

439: weaker than the outer two since the change in inter-electron interactions in 

440: going from the contacts to the quantum wire is likely to be much smaller than 

441: that in going from the contacts to the leads; also the change in geometry at 

442: the contact-quantum wire junction is likely to be much more adiabatic. Thus, 

443: we will from now on consider the junctions between the wire and the contacts 

444: and between the contacts and the leads as local barriers whose heights 

445: are determined by several factors, such as the nature of the inter-electron 

446: interaction and its screening, and the deviations from adiabaticity in the 

447: change in geometry in going from the reservoirs into the contacts or from the 

448: contacts into the quantum wire. To be general, we should take these four 

449: barriers to have different heights but it is very likely that any asymmetry 

450: between the left two and right two contacts will be small. Thus, we

451: can finally write the complete Hamiltonian for the quantum wire of 

452: spinless electrons, its contacts and its leads as 

453: \bea

454: H &=& (\int_{-\infty}^0+\int_{l+2d}^\infty ) ~

455: dx ~\frac{v_F}{2} [ \Pi(x)^2 + (\partial_x\phi(x))^2] \nonum \\ 

456: && +(\int_0^d+\int_{l+d}^{l+2d}) ~

457: dx ~\frac{v_C}{2K_C} [\Pi_C(x)^2 + (\partial_x\phi(x))^2] \nonum \\

458: && +\int_d^{l+d} ~dx ~\frac{v_W}{2K_W} [\Pi_W(x)^2 + 

459: (\partial_x\phi(x))^2] +V_{LC}\sin (2\sqpi\phi(0)) \nonum \\ 

460: && +V_{CW}\sin(2\sqpi\phi(d)+ 2k_Fd) 

461: +V_{WC}\sin(2\sqpi\phi(l+d)+ 2k_F(l+d)) \nonum \\ 

462: && +V_{CL}\sin(2\sqpi\phi(l+2d)+ 2k_F(l+2d))~,

463: \label{model}

464: \eea

465: where $\Pi_{C,W}(x) = (1/v_{C,W})\partial_t \phi(x)$.

466: Finally, it is worth commenting here that since our

467: model shows that a quantum wire with no disorder 

468: already has back-scattering junctions built into it, the notion of {\it ideal 

469: contact resistances} (which are 

470: seen in a study of this system using the Landauer-Buttiker formalism and arise 

471: from the {\it ideal} connection of the quantum wire to its reservoirs) which 

472: are universal in value, $h/2e^2$ to be precise, does not seem 

473: to hold true even for the so-called clean quantum wire with adiabatic 

474: junctions in the presence of inter-electron interactions within the quantum 

475: wire. We will show later that these junction barriers are likely to be weak 

476: when the lengths of the quantum wires are quite short or temperatures are

477: not very 

478: low, and that the junction barriers are likely to remain weak even after some 

479: small renormalization that might take place due to the electron-electron 

480: interactions in the quantum wire. Thus, the contact resistances between the 

481: wire and the reservoirs due to the junction barriers will be very nearly the 

482: universal value quoted above only for very short quantum wires (i.e., quantum 

483: point contacts) or when the 

484: temperatures are not very low. This is also observed in all the experiments 

485: till date \cite{tarucha,yacoby,thomas,liang}.

486: 

487: The generalization of the model to spinful fermions is straightforward. For 

488: completeness, we give below the Hamiltonian for spinful electrons in a quantum 

489: wire connected to external reservoirs through the contacts and junction 

490: barriers, 

491: \bea

492: H_{spin} &=& \sum_{i=\ua,\da}\Big[ (\int_{-\infty}^0+\int_{l+2d}^\infty) ~

493: dx ~\frac{v_{iF}}{2} [ \Pi_i(x)^2 + (\partial_x\phi_i(x))^2] \nonum \\ 

494: && +(\int_0^d+\int_{l+d}^{l+2d}) ~dx ~\frac{v_{iC}}{2K_{iC}}[\Pi_{iC}

495: (x)^2 + (\partial_x\phi_i(x))^2] \nonum \\

496: && + \int_d^{l+d} ~dx ~\frac{v_{iW}}{2K_{iW}} [\Pi_{iW}(x)^2 + 

497: (\partial_x\phi_i(x))^2] +V_{iLC}\sin (2\sqpi\phi_i(0)) \nonum \\ 

498: && + V_{iCW}\sin(2\sqpi\phi_i(d)+ 2k_{iF}d) 

499: +V_{iWC}\sin(2\sqpi\phi_i(l+d)+ 2k_{iF}(l+d))\nonum \\ 

500: && + V_{iCL}\sin(\sqpi\phi_i(l+2d)+ 2k_{iF}(l+2d)) \Big].

501: \label{modelspin}

502: \eea

503: Note that we have allowed for independent velocities and 

504: interaction strengths for the $\ua$ and $\da$ electrons. This

505: generality will be required when we study the model in the presence of

506: a magnetic field. Finally, let us note the fact that we will be taking into 

507: account only the outer two junction barriers (i.e., those at the junctions of 

508: the contacts and the leads) in all our subsequent calculations as these are 

509: likely to be the more significant junction barriers in the system as 

510: long as transport through fully open quantum wires is considered.

511: 

512: \section{Effective Actions}

513: 

514: In this section, the aim is to obtain an effective action 

515: in terms of the fields at the junction barriers for both spinless

516: and spinful electrons. We then analyze the symmetries of the 

517: effective action and obtain the resonance conditions.

518: 

519: \subsection{The case of spinless electrons}

520: 

521: In Sec. 2, it was shown that the screening out of the 

522: interactions in the 2DEG leads to a Hamiltonian with junction barriers given 

523: in Eq. (\ref{model}). The effective action for this model of spinless 

524: electrons can be written as

525: \beq

526: S = S_0 + S_{barrier} + S_{gate} ~,

527: \eeq

528: where we have defined each of the actions separately below.

529: \beq

530: S_0= \int d\tau ~[ \int_{-\infty}^0 dx {\cal L}_1 + \int_0^d dx {\cal

531: L}_2 + \int_d^{l+d} dx {\cal L}_3 + \int_{l+d}^{l+2d} dx {\cal L}_2

532: +\int_{l+2d}^{\infty} dx {\cal L}_1 ]~,

533: \label{s0}

534: \eeq

535: where

536: \beq

537: {\cal L}_1 ~=~ {\cal L} (\phi ; K_L , v_L ), ~~~ 

538: {\cal L}_2 ~=~ {\cal L} (\phi ; K_{C} , v_{C} ),

539: ~~~{\rm and} ~~~ {\cal L}_3 ~=~ {\cal L} (\phi ; K_{W} , v_{W}) ~,

540: \label{sol} 

541: \eeq

542: and we have defined ${\cal L} (\phi ; K,v) = (1/2Kv) (\pt \phi)^2 +(v/2K) 

543: (\px \phi)^2$, and used the imaginary time $\tau=it$ notation.

544: \bea 

545: S_{barrier} &=& \int d\tau ~\Lambda [ V_1 \cos (2\sqpi\phi_1) + V_1

546: \cos (2\sqpi\phi_4 +2k_FL) \nonum \\ 

547: && ~~~~~~~~~ +V_2\cos(2\sqpi\phi_2 + 2k_Fd) + V_2\cos(2\sqpi\phi_3 +

548: 2k_F(l+d)) ]~,

549: \eea 

550: where we have set $V_{LC} = V_{CL}=V_1 \Lambda$ and 

551: $V_{CW}=V_{WC} =V_2 \Lambda$ assuming 

552: left-right symmetric barriers ($V_1$ and $V_2$ are dimensionless), and have 

553: used $\phi(0,\tau) = \phi_1(\tau) 

554: \equiv \phi_1$, $\phi(d,\tau) = \phi_2(\tau) \equiv \phi_2$, $\phi(l+d,\tau)

555: = \phi_3(\tau) \equiv \phi_3$ and $\phi(L,\tau) = \phi_4(\tau) \equiv

556: \phi_4$. The total length of the wire is denoted by $L=l+2d$. 

557: We shall henceforth assume that $V_2\ll V_1$ and can be

558: dropped; as we have explained earlier, the inner two

559: barriers are likely to be weaker than the outer two barriers. 

560: We also include the coupling of the electrons in the wire 

561: to an external gate voltage $V_G$ given by 

562: \beq

563: S_{gate} ~=~ V_G \int d\tau \int_d^{l+d} dx \rho(x,\tau) ~

564: =~ V_G\int d\tau {[\phi_3 - \phi_2]\over \sqpi}.

565: \eeq

566: This coupling is necessary because it is the

567: gate voltage which controls the density of electrons in the wire,

568: which, in turn, controls the number of channels in the quantum

569: wire. Experimentally, an external voltage drop across the wire drives

570: the current through the wire, which is measured as a function of

571: the gate voltage or the density of electrons in the wire. 

572: 

573: Since the Luttinger liquid action is quadratic, the effective action 

574: can be obtained in terms of the fields $\phi_i$, $i=1...4$, by integrating

575: out all degrees of freedom except those at the positions of the

576: four junction barriers, following Ref. \cite{kane}. 

577: Using the (imaginary time) Fourier transform of the fields 

578: \beq

579: \phi_1(\tau) = \sum_{\on} e^{-i\on \tau} \tphi_{1n} (\on),

580: \quad \phi_2(\tau) = \sum_{\on} e^{-i\on \tau} \tphi_{2n} (\on) ~,

581: \eeq	

582: we explicitly obtain the $S_0$ part of the effective action; this is

583: presented in Appendix A. (The $\on$ are the Matsubara frequencies which are

584: quantized in multiples of the temperature as $\on = 2\pi n k_B T$).

585: In the high frequency limit, or, equivalently at high temperatures,

586: where $\on d/v_C, \on l/v_W \gg 1$, the effective action reduces to 

587: \beq

588: S_{0,eff,high} (\tphi_1,\tphi_2,\tphi_3,\tphi_4) 

589: ={K_L+K_C \over 2K_L K_C} \sum_{\on} |\on| (\tphi_1^2+\tphi_4^2)

590: +{K_W+K_C \over 2K_W K_C} \sum_{\on} |\on| (\tphi_2^2+\tphi_3^2).

591: \label{highfreq}

592: \eeq

593: In this limit, all the barriers are seen as the sum of individual 

594: barriers with no interference. In fact, if we integrate out the two

595: inner fields $\phi_2$ and $\phi_3$, we are just left with 

596: \beq

597: S_{0,eff,high}(\tphi_1,\tphi_4)

598: ={K_L+K_C \over 2K_L K_C} \sum_{\on} |\on| (\tphi_1^2+\tphi_4^2).

599: \eeq

600: The surprising point to note is that the effective interaction

601: strength $K_{eff} =K_L K_C /(K_L+K_C)$ depends only on the

602: interaction strengths in the contacts and in the leads (where there are

603: no interactions), and not on the interaction strength in the wire! 

604: Furthermore, since the gate voltage $V_G$ couples only to the inner fields 

605: $\phi_2$ and $\phi_3$ and these two fields are completely decoupled 

606: from the outer fields $\phi_1$ and $\phi_4$ in ${\cal L}_{0,eff,high} 

607: (\phi_1,\phi_2,\phi_3,\phi_4)$ above, integrating out $\phi_2$ and 

608: $\phi_3$ does not lead to any gate voltage term in the final effective 

609: action in this temperature regime.

610: 

611: Depending on whether $d\gg l$ or $l\gg d$, we can have two possible

612: scenarios of intermediate regimes, each with two 

613: crossovers. We can express all our lengths in terms of equivalent 

614: temperatures by defining $v_C/ d = k_B T_d$ and $v_W/l = k_B T_l$. 

615: So the high temperature limit defined above is just $T\gg T_d,T_l$. 

616: 

617: \begin{itemize}

618: 

619: \item{}

620: 

621: Let us first consider the quantum wire limit where $l\gg d$.

622: 

623: In the intermediate frequency (or temperature) regime of 

624: $T_l \ll T \ll T_d$, the action becomes

625: \bea

626: S_{0,eff,int} (\tphi_1,\tphi_2,\tphi_3,\tphi_4) &=&

627: {1\over 2K_L} \sum_{\on} |\on| (\tphi_1^2+\tphi_4^2) +

628: {1\over 2K_W} \sum_{\on} |\on| (\tphi_2^2+\tphi_3^2) \nonum \\

629: & & + {U_C \over 2} \sum_{\on} [ (\tphi_1- \tphi_2) ^2 

630: +(\tphi_3- \tphi_4)^2] + S_{gate} ,

631: \eea

632: where $U_C = v_C/(K_Cd)$ is an energy whose significance will 

633: become clear shortly. As the action is quadratic, we can integrate 

634: out $\tphi_2$ and $\tphi_3$ to be left with an action dependent only 

635: on $\tphi_1$ and $\tphi_4$ as given by

636: \bea

637: & & S_{0,eff,int} (\tphi_1,\tphi_4) \nonum \\

638: &=& {1\over 2K_L} \sum_{\on} |\on| (\tphi_1^2+\tphi_4^2) \nonum \\

639: & & + ~~{1\over 2K_W} \sum_{\on} |\on| 

640: [\frac{U_C^2}{A^2}(\tphi_1^2+\tphi_4^2) + 

641: \frac{2{\tilde V}_G U_C}{A^2}(\tphi_1-\tphi_4) ] \nonum \\

642: & & + ~~\frac{U_C}{2} \sum_{\on} \Bigl[ [ (\frac{U_C}{A} - 1)\tphi_1 + 

643: \frac{{\tilde V}_G}{A}]^2 + [ (\frac{U_C}{A} - 1)\tphi_4 - 

644: \frac{{\tilde V}_G}{A}]^2 \Bigr] \nonum \\

645: & & + \int d\tau \frac{{\tilde V}_G U_C}{A}(\phi_4 - \phi_1) ~,

646: \eea

647: where $A = U_C + |\on|/K_W$ and ${\tilde V}_G = V_G/\sqrt{\pi}$. We can 

648: approximate $A$ by $U_C$ which is justified in the 

649: intermediate regime as $T \ll T_d$ and $K_C \sim K_W$. 

650: Then we are finally left with the expression

651: \beq

652: S_{0,eff,int} (\tphi_1,\tphi_4) = {{K_L+K_W}\over{2K_LK_W} } \sum_{\on} |\on| 

653: (\tphi_1^2+\tphi_4^2) ~+~ \int d\tau {\tilde V}_G(\phi_4 - \phi_1) ~.

654: \label{intertemp}

655: \eeq

656: 

657: \item{}

658: 

659: Now, we consider the QPC limit where $d \gg l$.

660: 

661: In the regime where $T_d \ll T \ll T_l$, the action becomes

662: \bea

663: S_{0,eff,int} (\tphi_1,\tphi_2,\tphi_3,\tphi_4) &=&

664: {1\over 2K_L} \sum_{\on} |\on| (\tphi_1^2+\tphi_4^2) \nonum \\

665: & & + {1\over 2K_C} \sum_{\on} |\on| (\tphi_1^2+\tphi_2^2+

666: \tphi_3^2+\tphi_4^2) \nonum \\

667: & & + {U_W \over 2}\sum_{\on} (\tphi_2- \tphi_3)^2 + S_{gate} , 

668: \eea

669: where $U_W = v_W/(K_Wl)$ is again a frequency independent

670: energy. As before, we can integrate out $\tphi_2$ and $\tphi_3$ to be left 

671: with an action dependent only on $\tphi_1$ and $\tphi_4$ given by

672: \beq

673: S_{0,eff,int} (\tphi_1 ,\tphi_4 ) ~=~ S_{0,eff,high} (\tphi_1 ,\tphi_4 )~. 

674: \eeq

675: Thus there is no difference between the intermediate and high energy scales in

676: the QPC limit because the gate voltage is applied over too short a length to

677: affect the conductance even at intermediate temperatures.

678: 

679: \end{itemize}

680: 

681: Finally, in the low frequency limit where $\on \ll v_W/l , v_c/d$ 

682: (i.e., $T \ll T_d$ and $T \ll T_l$), $S_0$ reduces to 

683: \bea

684: S_{ 0,eff,low} (\tphi_1,\tphi_2,\tphi_3,\tphi_4) 

685: &=& {1\over 2K_L} \sum_{\on} |\on| (\tphi_1^2+\tphi_4^2) + {U_C\over 2} 

686: \sum_{\on} [ (\tphi_1- \tphi_2) ^2 +(\tphi_3- \tphi_4)^2] \nonum \\

687: && ~~~+{U_W\over 2} \sum_{\on} (\tphi_2- \tphi_3)^2 ~.

688: \eea

689: Since the action is still quadratic, it is possible to integrate out

690: the two inner fields $\tphi_2$ and $\tphi_3$ and get the effective action

691: wholly in terms of the $\tphi_1$ and $\tphi_4$ fields, remembering

692: however, to also include the gate voltage term which couples to the inner 

693: fields. After doing out, we are left with the full effective action as

694: \bea

695: S_{eff,low} (\phi_1,\phi_4) &=& S_{0,eff,low} + 

696: S_{gate}+ S_{barrier} \nonum \\

697: &=& {1\over 2K_L} \sum_{\on} |\on| (\tphi_1^2+\tphi_4^2) +

698: {U_C U_W \over 2(U_C+2U_W)}\sum_{\on} (\tphi_1-\tphi_4

699: -{{\tilde V}_G \over U_W})^2. \nonum \\

700: &&

701: \eea 

702: In this limit, the full action can be rewritten in terms of a 

703: ``current" field $\chi(\tau)$ and a ``charge" field $n(\tau)$

704: (and their Fourier transforms $\tilde \chi$ and $\tilde n$) defined as 

705: \beq

706: \chi(\tau) = {\phi_1 +\phi_4 \over 2 }+ {k_F L\over 2\sqpi}, \quad {\rm 

707: and} \quad n(\tau) = {\phi_4 - \phi_1 \over \sqpi} + {k_F L\over \pi} ~.

708: \eeq

709: The action is given by 

710: \bea

711: S_{eff,low} &=& S_0 + S_{barrier} +S_{gate} \nonum \\

712: &=& {1\over 2K_L} \sum_{\on} |\on| [(\tchi - {k_F 

713: L\over 2\sqpi})^2 + {\pi\over 4} (\tilde n -{k_F L\over \pi})^2] \nonum \\

714: && + \int d\tau [{U_{eff} \over 2} (n-n_0)^2 + V_1 \Lambda 

715: \cos (2\sqpi \chi) \cos (\pi n) ] ~,

716: \label{lowfreq}

717: \eea

718: where $n_0 = (2k_C d + k_W l)/\pi - V_G/(\pi^{3/2}U_W)$ and 

719: $U_{eff} =\pi U_C U_W / (U_C+2U_W)$. The derivation of the effective 

720: action in this limit follows the method outlined 

721: in Ref. \cite{kane}; however, their derivation was for a uniform wire with a 

722: single interaction parameter $K$, whereas we have three interaction 

723: parameters here. $K_W$ acts only within the quantum wire delimited by the two

724: contact regions, $K_C$ acts within the contact region, and $K_L =1$ outside 

725: the contact and wire region. 

726: The current field is interpreted as the number of particles transferred 

727: across the two barriers, and the charge field is the number of particles 

728: between the barriers. In the low frequency limit, the two barriers are 

729: clearly being seen as one coherent object with charge and current degrees of 

730: freedom. Since

731: in the limit of weak barriers, $V_1 \ll U_{eff}$, the action is minimized when 

732: $n=n_0$, we can integrate out the quadratic fluctuations of $n-n_0$ to obtain 

733: an effective action only in terms of the single variable $\chi$; we obtain

734: \beq

735: S_{low} = \int d\tau ~[~2V_1 \Lambda \cos (2\sqpi\chi) \cos (\pi n_0) 

736: -{2(V_1 \Lambda)^2 \over U_{eff}} (\pi \cos (2\sqpi \chi) \sin (\pi n_0))^2 

737: + \cdots ~]~.

738: \eeq

739: The first term in this effective action is precisely the

740: same term that is obtained for the impurity potential for 

741: a single barrier in terms of the variable $\chi$.

742: 

743: In the low frequency limit, from Eq. (\ref{lowfreq}), we see that the 

744: effective action contains extra terms due to the interference between the 

745: two barriers. It is easy to check that this effective action is invariant

746: under $\chi\rightarrow \chi+\sqrt{\pi}, n \rightarrow n$; this is 

747: the same symmetry which exists for a single barrier \cite{kane}, and it

748: corresponds to the transfer of a single electron across the two barriers, and 

749: hence in our model, from the left lead to the right lead. But when $n_0$ is 

750: precisely equal to a half-odd-integer, the action is also invariant 

751: under $\chi \rightarrow \chi+\sqpi/2, n\rightarrow 2n_0-n$. As explained in

752: Ref. \cite{kane}, this corresponds to the `transfer of half an electron

753: across the wire' accompanied by a change in the charge state of the 

754: wire. In the language of scattering, this corresponds to resonant tunneling

755: through a virtual state. Within the TLL theory, this is the explanation

756: of the Coulomb blockade phenomenon, which leads to steps or plateaus

757: in the current versus gate voltage for quantum dots.

758: 

759: \subsection{The case of spinful electrons}

760: 

761: The spinless electron model is expected to be valid for real systems

762: in the presence of strong magnetic fields which completely

763: polarizes all the electrons in a given channel. 

764: However, for a real system without magnetic 

765: field, or in the presence of weak magnetic fields which do not

766: polarize all the electrons, one has to study a model of electrons

767: with spin. We shall study such a model here. Its modification due

768: to the presence of a magnetic field will be studied in Sec. 5.

769: The basic action of the model is a straightforward extension of

770: the model for spinless fermions given in Eqs. (\ref{s0}) and (\ref{sol}), 

771: with $\phi_{\ua}$ denoting the spin up boson and $\phi_{\da}$ denoting

772: the spin down boson. However, since the Coulomb interaction couples the spin 

773: up and spin down fermions (for instance, remember the Hubbard term which is 

774: $U\sum_i n_{i\ua} n_{i\da}$), the Luttinger model is diagonal only in terms of 

775: the charge and spin fields $\phi_\rho =(\phi_{\ua} + \phi_{\da})/{\sqrt 2}$ 

776: and $\phi_\sigma =(\phi_{\ua} - \phi_{\da})/{\sqrt 2}$. In terms of 

777: these fields $S_0$ is given by

778: \beq

779: S_0 = \int dt ~[ \int_{-\infty}^0 dx {\cal L}_1 

780: + \int_0^d dx {\cal L}_2 + \int_d^{l+d} dx {\cal L}_3

781: + \int_{l+d}^{l+2d} dx {\cal L}_2 + \int_{l+2d}^{\infty} dx {\cal L}_1 ]~, 

782: \label{s0spin}

783: \eeq

784: where

785: \bea

786: {\cal L}_1 ~&=&~ {\cal L} (\phi_\rho ; K_{L\rho} , v_L ) ~+~ 

787: {\cal L} (\phi_\sigma ; K_{L\sigma} , v_L ) ~, \nonum \\

788: {\cal L}_2 ~&=&~ {\cal L} (\phi_\rho ; K_{C\rho} , v_{C\rho} ) ~+~ 

789: {\cal L} (\phi_\sigma ; K_{C\sigma} , v_{C\sigma} ) ~, \nonum \\

790: {\cal L}_3 ~&=&~ {\cal L} (\phi_\rho ; K_{W\rho} , v_{W\rho} ) ~

791: +~ {\cal L} (\phi_\sigma ; K_{W\sigma} , v_{W\sigma} ) ~,

792: \eea

793: with $\cal L$ defined as before, ${\cal L} (\phi ; K,v) = (1/2Kv) 

794: (\pt \phi)^2 +(v/2K) (\px \phi)^2$,

795: $K_{L\rho} = K_{L\sigma}=1$ are the interaction parameters in

796: the two external leads, and $K_{C/W\rho}$ and $K_{C/W\sigma}$ are the

797: interaction parameters in the contacts and wire respectively.

798: As for the spinless case, we include junction barrier terms of the form

799: \beq

800: S_{barrier} = \int d\tau \sum_{i=\da,\ua} ~V_i \Lambda ~[\cos (2\sqpi\phi_i(0,

801: \tau) ) +\cos (2\sqpi\phi_i(L,\tau) +2k_F L)] ,

802: \eeq

803: at the junctions of the contacts and the leads, and we assume that the

804: barriers at the junctions of the wire and the contact are weak and can

805: be ignored. The barrier action can be re-expressed in terms of the 

806: diagonal fields of the model (using $V_{\ua} = V_{\da} =V$) as 

807: \beq

808: S_{barrier} = V \Lambda~\int d\tau [\cos (\sqrt{2\pi}\phi_{1\rho}) \cos (

809: \sqrt{2\pi} \phi_{1\sigma}) + \cos (\sqrt{2\pi} \phi_{4\rho} +2k_F L)\cos (

810: \sqrt{2\pi} \phi_{4\sigma})] ~,

811: \eeq

812: where, as before, we define $\phi_\rho(0) = \phi_{1\rho}$ and $\phi_\rho(L) = 

813: \phi_{4\rho}$ and similarly for the $\phi_\sigma$ fields. The gate voltage 

814: only couples to the charge degree of freedom within the wire region as

815: \beq

816: S_{gate} = ~\frac{V_G}{\sqpi} ~\int d\tau 

817: (\phi_{3\rho} - \phi_{2\rho}) ~, 

818: \eeq

819: where $\phi_\rho(d)=\phi_{2\rho}$ and $\phi_\rho(l+d)=\phi_{3\rho}$ as 

820: before. So just as in the spinless case, we can integrate out all degrees 

821: of freedom except at $x=0,d,l$ and $L$ and obtain the effective

822: action. The full details of the effective action are spelt out in

823: Appendix B. By taking its high, intermediate and 

824: low frequency limits, we will be able to 

825: obtain conductance corrections just as we did for the spinless fermions.

826: 

827: In the high frequency limit where $\omega \gg v_{Ca} /d$ and $v_{Wa} /l$, the

828: two barriers are seen as decoupled barriers, with

829: \bea

830: && S_{0,eff,high} (\phi_{1,\rho/\sigma},\phi_{2,\rho/\sigma},

831: \phi_{3,\rho/\sigma},\phi_{4,\rho/\sigma}) \nonum \\

832: && \quad = \sum_{\on}

833: \vert \on \vert [ \frac{K_{L\rho} + K_{C\rho}}{2 K_{L\rho} K_{C\rho}} 

834: (\tphi_{1\rho}^2 + \tphi_{4 \rho}^2 ) + \frac{K_{C\rho} + K_{W\rho}}{2 

835: K_{C\rho} K_{W\rho}} (\tphi_{2\rho}^2 ~+~ \tphi_{3\rho}^2 ) \nonum \\

836: && ~~~~~~~~~~~~~ +{\rm ~~~similar ~terms ~with~} 

837: \rho \rightarrow \sigma ].

838: \eea

839: The fields $\phi_{2a}$ and $\phi_{3a}$ are completely decoupled

840: from the fields at $x=0$ and $L$ and can be integrated out yielding 

841: \bea

842: && S_{0,eff,high} (\phi_{1,\rho/\sigma},\phi_{4,\rho/\sigma}) 

843: \nonum \\

844: && \quad = {K_{L\rho}+K_{C\rho} \over 2K_{L\rho} K_{C\rho}} \sum_{\on} |\on| 

845: (\tphi_{1\rho}^2+\tphi_{4\rho}^2 ) + {K_{L\sigma}+ K_{C\sigma} \over 2 K_{L

846: \sigma} K_{C\sigma}} \sum_{\on} |\on| (\tphi_{1\sigma}^2+\tphi_{4\sigma}^2).

847: \eea

848: Just as for the spinless case, we see that the parameters of the wire

849: do not enter $K_{eff,\rho/\sigma} = K_{L\rho/\sigma} 

850: K_{C\rho/\sigma}/(K_{L\rho/\sigma}+K_{C\rho/\sigma})$. Nor does the

851: gate voltage affect the action. 

852: 

853: Just as in the spinless case, we have two possibilities for the intermediate

854: frequency regime, the QW limit or the QPC limit.

855: 

856: For the QW limit, we have $v_W/l \ll \on \ll v_C/d$, and we obtain 

857: \bea

858: &&S_{eff,int} (\tphi_{1,\rho/\sigma},\tphi_{2,\rho/\sigma},\tphi_{3,

859: \rho/\sigma}, \tphi_{4,\rho/\sigma}) \nonum \\ 

860: && \quad = {1\over 2K_{L\rho}} \sum_{\on} |\on| (\tphi_{1\rho}^2+ 

861: \tphi_{4\rho}^2) + {1\over 2K_{L\sigma}} \sum_{\on} |\on| (\tphi_{1\sigma}^2 

862: + \tphi_{4\sigma}^2) \nonum \\

863: && \quad + {1\over 2K_{W\rho}} \sum_{\on} |\on| (\tphi_{2\rho}^2+

864: \tphi_{3\rho}^2) + {1\over 2K_{W\sigma}} \sum_{\on} |\on| (\tphi_{2\sigma}^2 + 

865: \tphi_{3\sigma}^2) \nonum \\ 

866: && \quad + {U_{C\rho} \over 2} \sum_{\on} [ (\tphi_{1\rho}- \tphi_{2\rho}) ^2 

867: + (\tphi_{3\rho}- \tphi_{4\rho})^2] \nonum \\

868: && \quad + {U_{C\sigma} \over 2} \sum_{\on} [ (\tphi_{1\sigma}- 

869: \tphi_{2\sigma})^2 + (\tphi_{3\sigma}- \tphi_{4\sigma})^2] + S_{gate} ,

870: \eea

871: where $U_{C\rho,\sigma} = v_{C\rho,\sigma}/(K_{C\rho,\sigma}d)$ is the 

872: charging energy 

873: for the charge degrees of freedom. As the action is quadratic, we can 

874: integrate out the $\tphi_{2,\rho/\sigma}$ and $\tphi_{3,\rho/\sigma}$ spin 

875: and charge fields to be 

876: left with an action dependent only on the $\tphi_{1,\rho/\sigma}$ and 

877: $\tphi_{4,\rho/\sigma}$ spin and charge fields as given by

878: \bea

879: \hspace*{-1cm}S_{0,eff,int} (\phi_{1,\rho/\sigma}, \phi_{4,\rho/\sigma})

880: \hspace*{-0.3cm} &=& \frac{K_{L\rho}+K_{W\rho}}{2K_{L\rho}K_{W\rho}} 

881: \sum_{\on} |\on| (\tphi_{1\rho}^2+\tphi_{4\rho}^2) \nonum \\

882: && \hspace*{-0.5cm}+ \frac{K_{L\sigma}+K_{W\sigma}}{2K_{L\sigma}K_{W\sigma}} 

883: \sum_{\on} |\on| (\tphi_{1\sigma}^2+\tphi_{4\sigma}^2) + \int d\tau 

884: {\tilde V}_G (\phi_{4\rho} - \phi_{1\rho} ) ,

885: \eea

886: where we have approximated

887: $U_C + \on/(K_{W,\rho/\sigma})$ by $U_C$; this is justified in the 

888: intermediate regime as $T \ll v_{C,\rho/\sigma}/d$ and 

889: $K_{C,\rho/\sigma} \sim K_{W,\rho/\sigma}$. 

890: 

891: In the QPC limit, we have

892: \bea

893: &&S_{eff,int} (\tphi_{1,\rho/\sigma},\tphi_{2,\rho/\sigma},\tphi_{3,

894: \rho/\sigma}, \tphi_{4,\rho/\sigma}) \nonum \\ 

895: && \quad = {1\over 2K_{L\rho}} \sum_{\on} |\on| (\tphi_{1\rho}^2+

896: \tphi_{4\rho}^2) + {1\over 2K_{L\sigma}} \sum_{\on} |\on| (\tphi_{1\sigma}^2 

897: + \tphi_{4\sigma}^2) \nonum \\

898: && \quad + {1\over 2K_{C\rho}} \sum_{\on} |\on| 

899: (\tphi_{1\rho}^2 + \tphi_{2\rho}^2+\tphi_{3\rho}^2 +\tphi_{4\rho}^2)

900: + {1\over 2K_{C\sigma}} \sum_{\on} |\on| (\tphi_{1\sigma}^2 + 

901: \tphi_{2\sigma}^2 + \tphi_{3\sigma}^2 +\tphi_{4\sigma}^2) \nonum \\ 

902: && \quad + {U_{W\rho} \over 2}\sum_{\on} (\tphi_{2\rho}- \tphi_{3\rho}) ^2 

903: + S_{gate} ,

904: \eea

905: where $U_{W\rho} = v_{W\rho}/(K_{W\rho}l)$ is the charging energy 

906: for the charge degrees of freedom in the wire. 

907: As before, we may integrate out the inner degrees of freedom to find that

908: \beq

909: S_{0,eff,int} (\tphi_{1,\rho/\sigma}, \tphi_{4,\rho/\sigma}) ~=~ 

910: S_{0,eff,high} (\tphi_{1,\rho/\sigma}, \tphi_{4,\rho/\sigma})

911: \eeq

912: as expected.

913: 

914: Finally, in the low frequency limit $\on \ll v_C /d$ and $v_W /l$, 

915: as in the spinless case, the terms multiplying $1/K_{C\rho/\sigma}$ and 

916: $1/K_{W\rho/\sigma}$ in Eq. (\ref{spineff}) in Appendix B 

917: become constant `mass' terms. We get the full effective action as

918: \bea

919: &&S_{eff,low}

920: (\phi_{1,\rho/\sigma},\phi_{2,\rho/\sigma},\phi_{3,\rho/\sigma},

921: \phi_{4,\rho/\sigma}) \nonum \\ 

922: && \quad = \frac{1}{2 K_{L\rho}} \sum_{\on} |\on|

923: (\tphi_{1\rho}^2 + \tphi_{4\rho}^2) 

924: + \int d\tau [\frac{v_{C\rho}}{K_{C\rho}d} \{ (\phi_{1\rho} -\phi_{2

925: \rho})^2 + (\phi_{3\rho} - \phi_{4\rho})^2 \} \nonum \\

926: && \quad ~~ + \frac{v_{W\rho}}{K_{W\rho} } (\phi_{2\rho} - \phi_{3\rho})^2 

927: + {\rm similar ~terms ~with ~} ~\rho \rightarrow \sigma ] 

928: + \frac{eV_G}{\sqpi} \int d\tau [\phi_{3\rho} - \phi_{2\rho} ] \nonum \\

929: && \quad ~~ + V \Lambda \int d\tau [\cos (\sqrt{2\pi}\phi_{1\rho}) \cos (

930: \sqrt{2\pi} \phi_{1\sigma}) + \cos (\sqrt{2\pi}\phi_{4\rho} +2k_F L)\cos (

931: \sqrt{2\pi} \phi_{4\sigma})] ~. \nonum \\

932: &&

933: \label{lowfreqspin}

934: \eea

935: Just as we did in the spinless case, we now integrate out the fields 

936: at $x=d$ and $l+d$, in terms of which the above action is quadratic, to get 

937: \bea

938: &&S_{eff,low} (\phi_{1,\rho/\sigma},\phi_{4,\rho/\sigma})

939: \nonum \\

940: && \quad = {1\over 2K_{L\rho}}\sum_{\on} |\on| (\tphi_{1\rho}^2 +\tphi_{4

941: \rho}^2) + {U_{C\rho} U_{W\rho} \over 2(U_{C\rho}+2U_W{\rho})} \sum_{\on} 

942: (\tphi_{1\rho}-\tphi_{4\rho} -{{\tilde V}_G\over U_{W\rho}})^2 \nonum \\

943: && \quad ~~ + {1\over 2K_{L\sigma}}\sum_{\on} |\on| (\tphi_{1\sigma}^2 +

944: \tphi_{4\sigma}^2) + {U_{C\sigma} U_{W\sigma} \over 2(U_{C\sigma}+2

945: U_W{\sigma})} \sum_{\on} (\tphi_{1\sigma}-\tphi_{4\sigma})^2. 

946: \eea 

947: Here, we see that the effective mass terms are given by $U_{eff,\rho}

948: =\pi U_{C\rho} U_{W\rho} / (U_{C\rho}+2U_W{\rho})$ and

949: $U_{eff,\sigma} =\pi U_{C\sigma} U_{W\sigma} / (U_{C\sigma}+2U_{W\sigma})$ for 

950: the `charge' and `spin charge' fluctuations respectively. 

951: We denote the `charge on the quantum wire' 

952: fields as $n_\rho=\sqrt{2/\pi} (\phi_{1\rho} - \phi_{4\rho})$ and $n_\sigma=

953: \sqrt{2/\pi} (\phi_{1\sigma} - \phi_{4\sigma})$ respectively, and 

954: the `current' fields as $\chi_\rho = (\phi_{1\rho} + \phi_{4\rho})/\sqrt{2}, 

955: \chi_\sigma= (\phi_{1\sigma} + \phi_{4\sigma})/\sqrt{2}$ along with their 

956: appropriate Fourier transforms $\tchi_{\rho/\sigma}$ and 

957: ${\tilde n}_{\rho/\sigma}$ just as we did in the spinless case. The action 

958: then takes the following form,

959: \bea 

960: S_{eff,low} &=& {1\over 2K_{L\rho}} 

961: \sum_{\on} |\on| [|\tchi_{\rho} - {k_F d\over \sqpi}|^2 + {\pi\over 4} 

962: |\tilde n_{\rho} -{2k_F d\over \pi}|^2] + \int d\tau {U_{eff,\rho} \over 2} 

963: (n_{\rho} -n_{0\rho})^2 \nonum \\

964: && ~+ {1\over 2K_{L\sigma}} \sum_{\on}|\on| [|\tchi_{\sigma}|^2 + {\pi\over 4} 

965: |\tilde n_{\sigma}|^2] + \int d\tau {U_{eff,\sigma} \over 2} 

966: (n_{\sigma})^2 \nonum \\

967: && ~+ 2V \Lambda \int d\tau [\cos (\sqpi\chi_{\rho}) 

968: \cos ({\pi n_\rho\over 2}) \cos (

969: \sqpi\chi_{\sigma}) \cos({\pi n_\sigma\over 2}) \nonum \\ 

970: && ~~~~~~~~~~~~~~~~ +\sin (\sqpi\chi_{\rho}) \sin ({\pi n_\rho\over 2}) \sin (

971: \sqpi\chi_{\sigma}) \sin ({\pi n_\sigma\over 2}) ]~.

972: \label{reson}

973: \eea

974: We have used the fact that since it is only the $\rho$ field which couples

975: to the gate voltage and not the $\sigma$ fields, we only get 

976: $n_{0\rho} = (2k_C d + k_W l)/\pi - V_G/(\pi^{3/2}U_{W\rho})$ and 

977: $n_{0\sigma} =0$.

978: 

979: We now study the symmetries of the effective action to find out the 

980: possible resonances. As in the spinless fermion case, this effective action 

981: is invariant under the transformation $\chi_{\rho} \rightarrow \chi_{\rho} + 

982: \sqpi$ and $\chi_{\sigma} \rightarrow \chi_{\sigma} + \sqpi$, which

983: corresponds to the transfer of either an up electron or a down

984: electron through the two barriers. But besides this symmetry, there

985: are also some special gate voltages at which one can get resonance 

986: symmetries. This can happen when we adjust the gate voltage so as to make 

987: $n_{0\rho}$ an odd integer. In that case, $V_{eff}$ is invariant under

988: \bea

989: && n_\sigma \rightarrow -n_\sigma, ~~ n_\rho \rightarrow 2n_{0\rho} - n_\rho 

990: {\rm ~~in ~conjunction ~with} \nonum \\

991: && {\rm either} ~~ (i)~\chi_\rho\rightarrow \chi_{\rho} +\sqpi, ~~

992: \chi_\sigma \rightarrow \chi_\sigma ~~ {\rm or} ~~ (ii)~\chi_\sigma 

993: \rightarrow \chi_{\sigma} +\sqpi, ~~ \chi_\rho \rightarrow \chi_\rho ~. 

994: \eea

995: As explained in Ref. \cite{kane}, this resonance which occurs when $n_{0\rho}$ 

996: is tuned to be an odd integer, is called a Kondo resonance because it happens 

997: when two spin states in the island with $n_{\sigma}=\pm 1$ become degenerate.

998: 

999: The kind of resonance which was seen for spinless fermions when two

1000: charge states on the island becomes degenerate is harder to see for

1001: spinful fermions. Two charge states become degenerate when $n_{0\rho}$

1002: is tuned to be a half-odd-integer. But, in that case, the effective

1003: action in Eq. (\ref{reson}) does not have any extra `resonance symmetry' 

1004: unless $n_{0\sigma}$ (which we have set to be zero) is also tuned to be 

1005: a half-odd-integer. But non-zero $n_{0\sigma}$ is only possible 

1006: when there is an effective magnetic field or $SU(2)$ breaking field

1007: just over the quantum wire. This is because the Zeeman term is given by the

1008: Hamiltonian density

1009: \beq

1010: {\cal H}_{Zeeman} = -h(\partial_x\phi_\ua - \partial_x\phi_\da)

1011: = -\sqrt{2} h\partial_x\phi_\sigma ~,

1012: \eeq

1013: and it does not lead to any boundary terms as long as the magnetic field

1014: is felt through the full sample. However, 

1015: although in current experiments it is not possible to tune the $SU(2)$

1016: breaking to occur only between the two barriers, it could

1017: be possible in future experiments. Hence it is of interest to look

1018: for possible resonances in this case as well. We see that if one could arrange

1019: to tune both the gate voltage and the magnetic field (adjusted to be

1020: just over the quantum wire) so that $n_{0\rho}$ and $n_{0\sigma}$ are

1021: both half-odd-integers, the effective action in Eq. (\ref{reson}) is symmetric

1022: under $n_\rho \rightarrow 2n_{0\rho}-n_\rho$, $n_\sigma \rightarrow 

1023: 2n_{0\sigma}-n_\sigma$, $\chi_{\rho} \rightarrow \chi_{\rho} +\sqpi /2$, and

1024: $\chi_{\sigma} \rightarrow \chi_{\sigma} +\sqpi /2$. 

1025: This resonance is exactly analogous to the resonance that existed for

1026: spinless fermions and corresponds to hopping an electron from either

1027: of the leads to the wire. But since this requires the tuning of two

1028: parameters, it is a `higher' order resonance and will be more

1029: difficult to achieve experimentally.

1030:  

1031: In fact, if we allow for non-zero $n_{0\sigma}$, then the effective

1032: action also has the symmetry

1033: \bea

1034: && n_\rho \rightarrow -n_\rho, ~~ n_\sigma \rightarrow 2n_{0\sigma} - 

1035: n_\sigma ~~ {\rm in ~conjunction ~with} \nonum \\

1036: && {\rm either} ~~ (i) ~\chi_\rho\rightarrow \chi_{\rho} +\sqpi, ~~ 

1037: \chi_\sigma \rightarrow \chi_\sigma ~{\rm or~} ~~ (ii) ~\chi_\sigma 

1038: \rightarrow \chi_{\sigma} +\sqpi, ~~ \chi_\rho \rightarrow \chi_\rho ~,

1039: \eea

1040: when $n_{0\sigma}$ is an odd integer and $n_{0\rho}=0$. But this is hard

1041: to achieve, because one needs to tune the external gate voltage so as to 

1042: cancel the field due to the presence of all the other electrons within the 

1043: two barriers as well. Hence, this resonance will not be easy to see in 

1044: experiments. Moreover, it will show up in the spin conductance and not the 

1045: charge conductance.

1046: 

1047: In conclusion, we have studied in this section the effective actions

1048: of our model for both spinless and spinful fermions, and used them to

1049: obtain conductance corrections away from resonances (where the conduction

1050: is perfect) as a function of finite temperature and finite length of

1051: the wire. The same technique will again be used in Sec. 5, where it

1052: will be used to study the symmetries and obtain the conductance

1053: corrections of the quantum wire in the presence of a magnetic field.

1054: 

1055: \section{\bf Computation of the conductances}

1056: 

1057: In this section, we compute the conductances of our TLL quantum wire with 

1058: contacts, two semi-infinite Fermi liquid leads and two weak barriers at the 

1059: junctions of the contacts and the leads, for both spinless

1060: and spinful electrons, perturbatively in the barrier strength. We explicitly 

1061: derive an expression for the conductance to lowest order in 

1062: barrier strength (quadratic) in terms of the Green's functions of 

1063: the model. The RG flow of the barrier strengths has been incorporated 

1064: through a function $\chi(x,y)$. Thus, the behavior of

1065: the Green's functions in the different frequency regimes 

1066: determines the conductance corrections. The conductance corrections for a 

1067: simpler version of the model of the quantum wire (i.e., one in which the 

1068: quantum wire is directly connected to the Fermi leads through two weak 

1069: junction barriers) has already been studied by Safi

1070: and Schulz \cite{safilong}, who used a real time formulation

1071: and computed time-dependent Green's functions. The perturbative corrections 

1072: in the Kane-Fisher imaginary time formalism was also extended to the case of 

1073: finite length wires by Maslov \cite{maslovlong} and Furusaki and Nagaosa 

1074: \cite{furusaki}, who computed frequency dependent Green's functions. For our 

1075: model, with five distinct spatial regions with their boundaries, the real 

1076: time picture of TLL quasiparticle waves reflecting back and forth

1077: between the boundaries (as developed by Safi and Schulz \cite{safi}) is more 

1078: cumbersome; hence, we use the imaginary

1079: time formulation and compute frequency dependent Green's functions. 

1080: 

1081: \subsection{\bf The formulation of the conductance expressions}

1082: 

1083: The current through a clean quantum wire through which spinless 

1084: electrons are traveling can be found using the Kubo formula

1085: \beq 

1086: j(x) = \lim_{\omega\rightarrow 0}\int dy \sigma(x,y,\omega) E(y,\omega) ~,

1087: \eeq

1088: where $\sigma(x,y,\omega)$ is the non-local conductivity and is related to the 

1089: two-point Green's function $G(x,y,\omega)$ at finite frequency $\omega$ as

1090: \beq

1091: \sigma(x,y,\omega) = ~-i\frac{2 \omega e^2}{h} G(x,y,\omega) ~.

1092: \label{cond}

1093: \eeq

1094: For our model of the quantum wire, $G_{\bo}(x,y)$ has been computed in 

1095: Appendix D. Note that the real frequency $\omega$ is related to $\bar \omega$

1096: (used in the earlier sections) by the analytic continuation $\omega = i

1097: {\bar\omega} +\epsilon$. From Appendix D, we find that 

1098: $G_\bo(x,y) = K_L/(2|\bo|)$ + non-singular terms in $\bo$ in the limit 

1099: $\omega \rightarrow 0$ for our model. hence the dc conductance $g_0$ is given

1100: by

1101: \beq

1102: g_0 = \lim_{\omega\rightarrow 0}\sigma(x,y,\omega) = \frac{e^2}{h}~.

1103: \eeq

1104: This shows perfect dc conductance through the system 

1105: as in the earlier models without contacts\cite{safi,maslov}. 

1106: This result remains unchanged for the case of electrons with spin, 

1107: except for a multiplication of the conductance by a factor of two.

1108: 

1109: For a quantum wire in the presence of stationary impurities, 

1110: an explicit expression for the conductance can be derived to lowest

1111: (quadratic) order in the impurity strength from the partition function,

1112: using perturbation theory \cite{kane,safilong,maslovlong}. The 

1113: renormalization group (RG) equations for the barriers (discussed in detail in 

1114: subsection 4.4) imply that the barrier strengths grow under renormalization.

1115: However, it is only for very low temperatures or very long wire

1116: lengths that there will be considerable renormalization. 

1117: In real experimental setups, the length of the

1118: wire is in the range of micrometers and the temperatures in the

1119: range of a Kelvin; hence one does not expect much renormalization. 

1120: Hence, it is expected that the barrier strengths remain small 

1121: enough for perturbation theory to be applicable. We follow the methods 

1122: of Safi and Schulz \cite{safilong} and Maslov \cite{maslovlong}, who 

1123: derived explicitly a conductance expression for a non-translationally 

1124: invariant system and obtained 

1125: \beq

1126: g = g_0 K_L ( 1 - {\cal R} )~,

1127: \eeq

1128: where {\cal R} is the perturbative correction to second 

1129: order in the impurity strength. $\cal R$ is given by 

1130: \beq

1131: {\cal R} = g_0 K_L \sum_{m=1}^{\infty} m^2 c_m^2 {\cal R}^{(m)}~,

1132: \eeq

1133: where the $c_m$'s are the coefficients for the terms in the Haldane 

1134: representation of the fermionic density, and ${\cal R}^{(m)}$ is the 

1135: correction due to the back-scattering of {\it m} electrons given by

1136: \beq

1137: {\cal R}^{(m)} = \int\int dx dy V(x)V(y) \cos [ 2m(\xi (x) 

1138: - \xi (y))] \chi_m(x,y)~.

1139: \eeq

1140: In the above expression, $V(x)$ is the bare potential of the 

1141: impurities, $\xi$ is a phase factor which includes the 

1142: $k_F x$ factor coming from the 

1143: back-scattering process and other factors which arise due to the removal 

1144: of the forward scattering terms from the Hamiltonian by shifts in the 

1145: bosonic field $\phi$, and $\chi_m(x,y)$ is a factor which 

1146: incorporates the renormalization group (RG) flows of the barrier strengths.

1147: In general, it is given by a two-point correlation function defined as 

1148: \beq

1149: \chi_m(x,y) = \frac{1}{T^2} e^{-2m^2G_0(x,x,\tau_0) - 2m^2G_0(y,y,\tau_0)}

1150: \int_0^{\infty} dt e^{4m^2G_0(x,y,it+\pi/2)}~,

1151: \eeq 

1152: $G_0 (x,y,it)$ is the two-point Green's function for a clean quantum wire. 

1153: Here, the Green's functions are in terms of the imaginary time $\tau =it$. 

1154: $\tau_0 \sim 1/\Lambda$ is the inverse of the high energy cutoff.

1155: In a later subsection, we show how the one-point function 

1156: $\chi_m(x,x)$ can be obtained directly from the RG equation for the barriers.

1157: 

1158: For our system,

1159: we shall instead compute the two-point Green's function in terms of $\bar

1160: \omega$, in terms of which, the correction to the conductance is given by 

1161: (specializing to the case $m=1$) 

1162: \bea 

1163: R^{(1)} &=& {\rm lim}_{\omega \rightarrow 0} 

1164: \frac{\omega}{(g_0 K_L)^2} \int dx' \int dy' G_0(x,x',\omega ) 

1165: G_0(y,y',\omega)\cos [2(\xi (x') - \xi (y'))] \times \nonum \\

1166: && \quad \quad \quad \quad ~~~~~~~~~~~~~~~~~~~~ V(x') V(y') 

1167: Im(F(x',y',\omega)-F(x',y',0)) \nonum \\

1168: &=& \int dx' \int dy' \cos [2(\xi (x') - \xi (y'))] V(x') V(y')

1169: {\rm lim}_{\omega \rightarrow 0} {dF(x',y',\omega) \over d \omega} ,

1170: \eea

1171: where 

1172: \bea

1173: && \hspace*{-1cm} F(x',y',\omega) \nonum \\

1174: && \hspace*{-1cm}= \int_{-\infty}^{\infty} dt e^{i\omega t} \exp

1175: ( - 2\pi \sum_{\on'} [G_0(x',x',\on') + G_0(y',y', \on') - 2 G_0(x',y',\on')

1176: \cos (\on' \tau)]).

1177: \eea

1178: Note that the Green's functions in the prefactors of $R^{(1)}$ are dependent

1179: on the external driving frequency, but the Green's functions in

1180: the exponential depend only on the Matsubara frequencies $\on'$, and not

1181: on the external driving frequency $\omega$ or its analytic continuation

1182: $\bo$. The sum over the Matsubara frequencies are cutoff at the low energy 

1183: end by $ {\bar\omega}'_{n=1} \sim k_B T$ and at the upper end 

1184: by the high energy cutoff $\Lambda$. In evaluating $R^{(1)}$, we will

1185: approximate $\sum_{\on'}$ by $\int d\bo' /(2\pi)$ which is reasonable

1186: since we always assume that the temperature $T$ is much smaller than the

1187: cutoff $\Lambda$.

1188: 

1189: \subsection{Results for the Quantum Wire}

1190: 

1191: We will concentrate here on calculating the conductance of a quantum wire 

1192: system in which the length of the quantum wire $l$ is much greater than 

1193: the length of the contact regions $d$. Also, we will finally be interested 

1194: in studying the effects of the junction barriers placed at the two 

1195: lead-contact junctions (as explained earlier). Hence, for our model

1196: \beq

1197: V(x) = V_1 \Lambda \delta(x) + V_2 \Lambda \delta(x-L).

1198: \eeq

1199: For this potential, we can obtain the expression for the conductance 

1200: corrections

1201: as

1202: \bea

1203: R^{(1)} &=& \Lambda^{2} {\rm lim}_{\omega \rightarrow 0} 

1204: [V_1^2 Im {dF(0,0,\omega)

1205: \over d\omega} + V_2^2 Im {dF(L,L,\omega)\over d\omega} \nonum \\

1206: && \quad \quad ~~~~~+ 2V_1 V_2 Im {dF(0,L,\omega)\over d\omega}

1207: \cos(\xi(0)-\xi(L))]. 

1208: \label{r1}

1209: \eea

1210: 

1211: \begin{itemize}

1212: 

1213: \item{} Spinless electrons

1214: 

1215: Now, the expression for the one-point Green's function (in frequency space) 

1216: for a barrier placed inside 

1217: the contact region on the left of the QW and at a distance $a$ from the left 

1218: lead-contact junction (which is taken to be the origin, giving the hierarchy 

1219: of length scales $a \ll d \ll l$) can be easily obtained from Appendix D.

1220: It is given by

1221: \beq

1222: G_{\bar{\omega}} (x=a,y=a) \simeq \frac{K}{2\vert\bar{\omega}\vert} ~,

1223: \eeq

1224: where

1225: \begin{displaymath}

1226: K = \left\{ \begin{array}{ll}

1227: K_C & \textrm{for $\vert\bar{\omega}\vert \gg v_C/a$}\\ \\

1228: \frac{2K_LK_C}{K_L+K_C} & \textrm{for $v_C/d \ll \vert\bar{\omega}\vert 

1229: \ll v_C/a$}\\ \\

1230: \frac{2K_LK_W}{K_L+K_W} & \textrm{for $v_W/l \ll \vert\bar{\omega}\vert 

1231: \ll v_C/d$}\\ \\

1232: K_L & \textrm{for $\vert\bar{\omega}\vert \ll v_W/l$ ~.}

1233: \end{array} \right.

1234: \end{displaymath}

1235: For our model with a barrier at the left lead-contact junction, $a=0$ and 

1236: the first frequency regime does not exist.

1237: 

1238: The two-point Green's function $G(x,y)$ 

1239: for $y$ in the left contact region and $x$ anywhere is given in 

1240: Appendix D. By setting $y=0$ (i.e., at the first barrier) and $x=L=l+2d$

1241: (i.e., at the second barrier) we 

1242: obtain the conductances in the different frequency regimes given by

1243: \bea

1244: G &=& {2K_C(1+\gamma_1)^2 \over (2+K_C/K_W + K_W/K_C)} {\exp[-|\bo|

1245: ({2d\over v_C} +{l\over v_W})] \over |\bo|} \quad {\rm for}\quad 

1246: |\bo| \gg v_W/l \gg v_C/d 

1247: \nonum \\

1248: &=& {2K_C(1+\gamma_1)^2 \over (2+K_C/K_W + K_W/K_C)} {\exp[-|\bo|

1249: ({2d\over v_C})] \over |\bo|} ~~\quad \quad \quad {\rm for} \quad 

1250: v_W/l \gg |\bo| \gg v_C/d \nonum \\

1251: &=& {K_L \over 2|\bo|} \quad \quad {\rm for} \quad v_C/d,v_W/l \gg |\bo| ,

1252: \label{twopoint}

1253: \eea

1254: where $\gamma_1 = {K_L-K_C\over K_L+K_C}$. 

1255: Thus, we see that $G(x,y, \bo')$ decays 

1256: exponentially to zero except at the lowest frequency

1257: regime where $G(x,y,\bo') = G(x,x,\bo') = G(y,y,\bo')$.

1258: 

1259: To obtain the conductance corrections, we

1260: use the above Green's functions to compute $F(x',y',\bo')$ in each of these 

1261: frequency regimes. For the high frequency regime, it is simply given by

1262: \bea

1263: F(x',y',\omega) &=& \int_{-\infty}^\infty dt 

1264: e^{i\omega t} \exp [-2\pi \int_T^\Lambda {d\bo'\over 2\pi} 

1265: ({K_{eff} \over |\bo'|} - 2 G(x',y',\bo') \cos \bo' \tau )] \nonum \\

1266: &=& \int_{-\infty}^\infty {1\over T} dz e^{i\omega z/T}

1267: ({T\over \Lambda})^{2K_{eff}} \exp [-2\pi \int_T^\Lambda

1268: {d\bo'\over 2\pi} 2 G(x',y',\bo') \cos \bo' \tau] , \nonum \\

1269: &&

1270: \eea

1271: where $K_{eff} = K_L K_C/(K_L + K_C)$, and 

1272: in the second line, we have scaled $t$ by $T$, i.e., we

1273: have used $t=z/T$ to write the integral in terms of dimensionless

1274: variables so that the temperature power-laws can be made explicit.

1275: When $x'\ne y'$, $G(x',y', \bo') \rightarrow 0$, so that one can check that

1276: $lim_{\omega \rightarrow 0}Im {dF\over d\omega}$ also tends to zero.

1277: This means that the cross-term in Eq. (\ref{r1}) does not contribute.

1278: For each of the terms involving just one barrier, we find that 

1279: \beq

1280: lim_{\omega \rightarrow 0}Im {dF\over d\omega} = 

1281: {1\over T^2}({T\over \Lambda})^{2K_{eff}} \exp [- \int_T^\Lambda

1282: {d\bo'} {K_{eff}\over |\bo'|} \cos \bo' \tau].

1283: \eeq 

1284: Hence, we obtain the following answer for the conductance correction

1285: for high temperatures $T\gg T_d \equiv v_C/(k_B d)$, 

1286: \beq

1287: g = g_0 K_L [ 1 - c_1 ({T\over \Lambda})^{2(K_{eff} - 1)}(|V_1|^2+ |V_2|^2)],

1288: \label{hightemp}

1289: \eeq

1290: where $c_1$ is a dimensionful constant 

1291: dependent on factors like the contact quasiparticle velocity $v_C$, but is

1292: independent of the gate voltage $V_G$.

1293:  

1294: For intermediate temperatures 

1295: where $T_l \equiv v_W/(k_B l) \ll T \ll T_d$ the calculation is very

1296: similar to that performed for the high frequency case, except that the 

1297: integral over the Matsubara frequencies is now split into two regions

1298: \bea

1299: F(x',y',\omega) &=& \int_{-\infty}^\infty dt 

1300: e^{i\omega t} \exp [-2\pi (\int_T^{T_d} + \int_{T_d}^\Lambda)

1301: {d\bo'\over 2\pi} {K \over |\bo'|} ]~ \nonum \\

1302: && \quad \quad \quad \exp

1303: [2\pi \int_T^\Lambda {d\bo'\over 2\pi} 2 G(x',y', \bo') \cos \bo' \tau] ~.

1304: \eea

1305: The rest of the calculations go through as above, and we find that

1306: the conductance expression is 

1307: \beq

1308: g = g_0 K_L [ 1 - c_2 ({T_d\over \Lambda})^{2(K_{eff} - 1)}

1309: ({T\over T_d})^{2({\tilde K}_{eff}-1)} (|V_1|^2 + |V_2|^2) ],

1310: \eeq

1311: where ${\tilde K}_{eff} = K_L K_W/(K_L + K_W)$ and $c_2$ is a dimensionful 

1312: constant which is dependent on $V_G$. 

1313: 

1314: Finally, for very low temperatures $T \ll T_l$, the sum over 

1315: Matsubara frequencies split into three regions so that we have

1316: \bea

1317: F(x',y',\omega) &=& \int_{-\infty}^\infty dt 

1318: e^{i\omega t} \exp [-2\pi (\int_T^{T_l} + \int_{T_l}^{T_d}

1319: +\int_{T_d}^\Lambda) {d\bo'\over 2\pi} {K \over |\bo'|} ] ~ \nonum \\

1320: && \quad \quad \quad \exp

1321: [2\pi \int_T^\Lambda {d\bo'\over 2\pi} 2 G(x',y', \bo') \cos \bo' \tau] ~.

1322: \eea

1323: Furthermore, in this regime, the cross-term does not vanish; in

1324: fact, for $x' \ne y'$, we have $G(x',y',\bo) = G(x',x',\bo)=G(y',y',\bo)$,

1325: so that $F(x',y',\omega) =F(x',x',\omega) =F(y',y',\omega)$.

1326: The contribution of the cross term is hence identical to that

1327: of the terms due to a single barrier. Hence, we obtain the

1328: corrections to the conductance as

1329: \beq

1330: g = g_0 K_L [ 1 - c_3 ({T\over T_l})^{2(K_L - 1)} ({T_d \over 

1331: \Lambda})^{2(K_{eff} -1)} ({T_l\over T_d})^{2({\tilde K}_{eff} - 1)}

1332: |V_1 + V_2|^2 ] ~,

1333: \eeq

1334: where $c_3$ is a dimensionful constant similar in nature to $c_2$; thus the 

1335: two barriers are seen coherently.

1336: 

1337: Note that the power-laws come purely from the one-point Green's functions,

1338: whereas the phase coherence between the barriers is determined by the

1339: behavior of the two-point correlation function. At high or intermediate 

1340: frequencies, $lim_{\omega \rightarrow 0} Im {dF (x',y',\omega) \over 

1341: d\omega}$ tends to zero for $x'\ne y'$ leading to the lack of phase coherence 

1342: between the two barriers. At very high temperatures, the 

1343: interaction parameters of the contact region $K_C$ and the lead region $K_L$ 

1344: controls the renormalization of a barrier in the contact region. 

1345: As the temperature is lowered, the phase coherence length of the electronic 

1346: excitations increases, and the renormalization exponent makes a crossover to 

1347: a combination of the interaction parameters of the contact and QW, and finally 

1348: to that of the lead alone at the lowest temperature regime. The lowest 

1349: temperature regime is also the one in which resonant transport through both 

1350: the lead-contact junction barriers can take place as phase coherence over the 

1351: entire system is achieved at these temperatures.

1352: 

1353: \item{} Electrons with spin

1354: 

1355: The above expressions were given for a 

1356: model of the QW system but for spinless electrons. Let us now see what the 

1357: conductance expressions are for electrons with spin. These expressions can 

1358: be derived in the same way as for spinless electrons by using the appropriate

1359: Green's functions for spin and charge fields. This

1360: gives us for the high temperature regime of $T_d \ll T $

1361: \beq

1362: g = 2 g_0 K_L [ 1 - c_4 ({T \over \Lambda})^{2(K_{eff} - 1)} 

1363: (|V(0)|^2 + |V(l+2d)|^2) ],

1364: \eeq

1365: where now $K_{eff} = K_L K_{C\rho}/(K_L + K_{C\rho}) + K_L K_{C\sigma}/(K_L 

1366: + K_{C\sigma})$, and $c_4$ is a dimensionful constant much like $c_1$ for the 

1367: spinless case (i.e., dependent on the contact charge velocity $v_{C,\rho}$ but 

1368: independent of the gate voltage $V_G$). For the intermediate temperature range 

1369: $T_l \ll T \ll T_d$, we find 

1370: \beq

1371: g = 2 g_0 K_L [ 1 - c_5 ({T_d \over \Lambda})^{2(K_{eff} - 1)}

1372: ({T \over T_d})^{2({\tilde K}_{eff}-1)} (|V(0)|^2 + |V(l+2d)|^2) ] ~,

1373: \eeq

1374: where ${\tilde K}_{eff} = K_L K_{W\rho}/(K_L + K_{W\rho}) + 

1375: K_L K_{W\sigma}/(K_L + K_{W\sigma})$, and $c_5$ is a constant similar to 

1376: $c_2$ for the spinless case (i.e., dependent on $V_G$). Finally, for the low 

1377: temperature regime $T \ll T_l$, we obtain

1378: \beq

1379: g = 2 g_0 K_L [ 1 - c_6 ({T \over T_l})^{2(K_L - 1)} ({T_d \over \Lambda})^{2

1380: (K_{eff} -1)} ({T_l \over T_d})^{2({\tilde K}_{eff} - 1)}

1381: |V(0) + V(l+2d)|^2 ] ~,

1382: \eeq

1383: where $c_6$ is similar in nature to $c_3$ for the spinless case.

1384: 

1385: \end{itemize}

1386: 

1387: \subsection{Results for the Quantum Point Contact}

1388: 

1389: The Quantum Point Contact (QPC) is simply a quantum wire system in which the 

1390: length of the quantum wire region $l \sim 0.2 - 0.5 \mu m$ (i.e., the region 

1391: undergoing the constriction due to the application of the gate voltage) 

1392: is much reduced in comparison to typical lengths for a quantum wire 

1393: $l \sim 2 - 20 \mu m$. Thus, in our model of the quantum wire system, we can 

1394: reach the QPC by studying the limit when the contact region length $d$ is much 

1395: greater than the wire length $l$. Let us then study the effects of 

1396: barriers/impurities placed in the contact and wire region of the QPC.

1397: 

1398: \begin{itemize}

1399: 

1400: \item{} Spinless electrons

1401: 

1402: In order to study the effect of a weak barrier placed in the contact 

1403: region such that its distance $a$ from the left lead-contact junction falls 

1404: in the hierarchy of $a \ll l \ll d$, we again start by computing the one-point 

1405: Green's function for such an impurity. We find that

1406: \beq

1407: G_{\bar{\omega}} (x=a,y=a) \simeq \frac{K}{2\vert\bar{\omega}\vert},

1408: \eeq

1409: where

1410: \begin{displaymath}

1411: K = \left\{ \begin{array}{ll}

1412: K_C & \textrm{for $\vert\bar{\omega}\vert \gg v_C/a$}\\ \\

1413: \frac{2K_LK_C}{K_L+K_C} & \textrm{for $v_W/l \ll \vert\bar{\omega}\vert 

1414: \ll v_C/a$}\\ \\

1415: \frac{2K_LK_C}{K_L+K_C} & \textrm{for $v_C/d \ll \vert\bar{\omega}\vert 

1416: \ll v_W/l$}\\ \\

1417: K_L & \textrm{for $\vert\bar{\omega}\vert \ll v_C/d$ ~.}

1418: \end{array} \right. 

1419: \end{displaymath}

1420: 

1421: As before, for the two-point function, we find that the high and low

1422: frequency limits are the same as that given in Eq. (\ref{twopoint}) for the 

1423: QW, but for $T_d \ll T \ll T_l$, the answer turns out to be

1424: the same as in the high frequency limit. This is similar

1425: to what one sees for the one-point Green's functions above as well.

1426: So, without giving any further derivations,

1427: we directly quote the expressions for the conductance corrections.

1428: In the high and intermediate frequency regimes, the conductance is given by

1429: \beq

1430: g = g_0 K_L [ 1 - c_i 

1431: ({T\over \Lambda})^{2(K_{eff} - 1)} (|V_1|^2+ |V_2|^2) ],

1432: \eeq

1433: where $i=4,5$ allows for the constant to be different in the high and 

1434: intermediate frequency regimes. For the low frequency regime, we get

1435: \beq

1436: g = g_0 K_L [ 1 - c_6

1437: ({T\over T_d})^{2(K_L - 1)} ({T_d\over \Lambda})^{2(K_{eff} - 1)} 

1438: (|V_1 + V_2|^2) ],

1439: \eeq

1440: It is clear from the above expressions 

1441: that the contributions of barriers in the contacts of a QPC are always 

1442: going to be independent of the gate voltage $V_G$ as the QPC interaction 

1443: parameter $K_W$ does not enter anywhere. Thus, such an impurity would always 

1444: lead to a flat and channel independent renormalized conductance. 

1445: It should be noted that we have found from a similar calculation that even 

1446: for an impurity placed deep inside the contact (i.e., with the hierarchy of 

1447: $l \ll a \ll d$), the above conclusions still remain true; this is because 

1448: the only change that takes place is that $K = K_C$ (rather than the 

1449: combination of $K_L$ and $K_C$ found earlier) for the regime of $v_W/l \ll 

1450: \vert\bar{\omega}\vert \ll v_C/a$.

1451:  

1452: Finally, let us study the effect of an impurity placed inside the QPC itself. 

1453: We find the one-point Green's function for such a case (with the hierarchy of 

1454: $l \ll d < a$) to be

1455: \beq

1456: G_{\bar{\omega}} (x=a,y=a) \simeq \frac{K}{2\vert\bar{\omega}\vert},

1457: \eeq

1458: where

1459: \begin{displaymath}

1460: K = \left\{ \begin{array}{ll}

1461: K_W & \textrm{for $\vert\bar{\omega}\vert \gg v_W/l$}\\ \\

1462: K_C & \textrm{for $v_C/d \ll \vert\bar{\omega}\vert 

1463: \ll v_W/l$}\\ \\

1464: K_L & \textrm{for $\vert\bar{\omega}\vert \ll v_C/a$ ~.}

1465: \end{array} \right. 

1466: \end{displaymath}

1467: We have here only three frequency regimes as the regime of 

1468: $v_C/d \ll \vert\bar{\omega}\vert \ll v_W/l$ cannot be taken sensibly within 

1469: the given hierarchy of length scales. This shows again that the effect of an 

1470: impurity placed within the QPC will always be dependent on the gate voltage 

1471: $V_G$, and can never lead to flat and channel independent renormalizations of 

1472: the conductance. 

1473: 

1474: \item{} Electrons with spin

1475: 

1476: The generalization to spinful electrons can be obtained just

1477: as was done for quantum wires with the appropriate substitutions.

1478: 

1479: \end{itemize} 

1480: 

1481: \subsection {Evaluation of the conductances from the effective

1482: actions using the RG equations}

1483: 

1484: Here, we note that the above results for the conductances

1485: could have been anticipated by computing the RG equation for the impurity 

1486: potentials using the effective actions calculated in Sec. 3.

1487: 

1488: The conductance is governed by the renormalized barrier potentials at the two 

1489: junctions. Since the interaction is repulsive, the barrier potentials

1490: are expected to grow as a function of the frequency cutoff. This is

1491: what leads to the result that any impurity potential,

1492: however small, eventually cuts the wire; in the zero temperature

1493: limit, there is no transmission at all \cite{kane}. However, 

1494: at a finite temperature $T$, finite wire length $l$ or finite contact

1495: length $d$, the growth is cutoff by either $T$, $v_W/L$ or $v_C/d$.

1496: In fact, since the energy scales in the problem are the 

1497: temperature $k_BT$, the high frequency cutoff $\Lambda$ and those related 

1498: to the contact length $k_BT_d= v_C/d$ and 

1499: the wire length $k_BT_l= v_W/l$, we can see that there will 

1500: exist two energy scale crossovers in the system --- one from $T/\Lambda$ 

1501: to $T/T_d$ and the other from $T/T_d$ to $T/T_l$ for $d \ll l$ (the QW 

1502: limit), or from $T/\Lambda$ to $T/T_l$ and then from $T/T_l$ to $T/T_d$ for 

1503: $l \ll d$ (the QPC limit). 

1504: 

1505: In fact, an explicit RG calculation of either of the individual barrier

1506: strengths in the high, intermediate and low frequency regimes 

1507: simply involves computing the dimension

1508: of $\cos (2\sqpi\phi_1)$ or $\cos (2\sqpi\phi_4)$ (which turn out to be 

1509: the same) using those respective actions. For example, for the high 

1510: frequency effective action given in Eq. (\ref{highfreq}), the RG 

1511: equation for a single barrier is given by

1512: \beq

1513: {dV_1\over d\lambda} = (1 - {2K_C K_L\over K_C+K_L}) V_1 \equiv 

1514: (1 - K_{eff})V_1 ~,

1515: \eeq

1516: where $\lambda={\rm ln} \frac{\Lambda(\lambda)}{\Lambda}$.

1517: Using this, we can get the renormalized barrier strength to be

1518: \beq

1519: V_1^{ren} = V_1 ({T\over \Lambda})^{K_{eff}-1}

1520: \eeq

1521: in the high frequency regime $T\gg T_d$, where we have used $T$ to 

1522: cutoff the RG flow (which begins from $\Lambda$). From this, we infer that

1523: to quadratic order, the $T$ dependence of the conductance 

1524: corrections is given by

1525: \beq

1526: g = {e^2\over h} [1 - {\tilde c}_1V_1^2 (\frac{T}{\Lambda})^{2(K_{eff}-1)}]

1527: \quad {\rm for} \quad

1528: T\gg T_d ~,

1529: \label{offres}

1530: \eeq 

1531: where ${\tilde c}_1$ is a dimensionful constant like $c_1$ defined 

1532: earlier containing factors like $v_C$ but 

1533: is, most importantly, independent of the gate voltage $V_G$.

1534: Comparing with Eq. (\ref{hightemp}), we see that if we include the 

1535: subtraction due to two barriers, the expressions are identical.

1536: 

1537: In the intermediate regime of $T_l \ll T \ll T_d$, the RG equation for the 

1538: same barrier now becomes

1539: \beq

1540: {dV_1\over d\lambda} = (1 - {2K_C K_W\over K_W+K_L}) V_1 \equiv

1541: (1 - {\tilde K}_{eff})V_1 ~.

1542: \eeq

1543: using the effective action in Eq. (\ref{intertemp}).

1544: At the same time, the appearance of the energy scale $v_C/d$ (through 

1545: $U_C$) in the effective action in this temperature regime and the taking 

1546: of the approximation $\on \ll U_C$ means that $v_C/d$ 

1547: has replaced $\Lambda$ as the high energy cutoff in the expression for 

1548: the $T$ dependence of the conductance correction. The influence of 

1549: those degrees of freedom whose energies lie between $v_C/d$ and $\Lambda$ 

1550: can be taken into account by noting that they will contribute a 

1551: factor of $(T_d/\Lambda)^{2(K_{eff}-1)}$; this is because 

1552: these degrees of freedom have been integrated away during the RG 

1553: procedure, and there must be continuity between the 

1554: conductance expressions for $T \gg T_d$ and $T_l \ll T \ll T_d$ at 

1555: $T = T_d$. Thus, we get the conductance expression in this regime as

1556: \beq

1557: g = {e^2\over h} [1 - {\tilde c}_2V_1^2 (\frac{T_d}{\Lambda})^{2(K_{eff}-

1558: 1)}) (\frac{T}{T_d})^{2({\tilde K}_{eff}-1)}]

1559: \quad {\rm for} \quad T_l\ll T\ll T_d ~,

1560: \label{intcond}

1561: \eeq

1562: where ${\tilde c}_2$ is a constant similar in nature to ${\tilde c}_1$, but it 

1563: can depend on $v_W$ and is hence dependent on the gate voltage $V_G$. Thus the 

1564: conductance is no longer independent of $V_G$. This again is the same as the 

1565: expression obtained by the explicit computation of the conductance.

1566: 

1567: Finally, in the low temperature limit, 

1568: we recognize the fact that there is phase coherence over the distance 

1569: between the two barriers; this follows from

1570: the low frequency effective action which has cross terms between

1571: the fields at the two barriers. This is what leads to resonant transmission. 

1572: To compute the conductance corrections away from resonance in this limit, 

1573: we note the following. Since the resonance occurs precisely when the 

1574: $2k_F$ component of the barrier term goes to zero, the relevant

1575: term away from this resonance is precisely the back-scattering potential

1576: $V\cos (2\sqpi \chi)$. Computing the dimension of this operator gives us the 

1577: RG equation for our barriers in this temperature regime as

1578: \beq

1579: {dV_1\over d\lambda} = (1 - K_L) V_1 .

1580: \eeq

1581: This makes the $T$ dependence of the conductance correction clear. Again, 

1582: the appearance of the energy scale $v_W/l$ (through $U_W$) and 

1583: the approximation $\on \ll U_W$ indicate 

1584: that $v_W/l$ has now replaced $v_C/d$ as the high energy cutoff in the 

1585: expression for the $T$ dependence of the conductance correction. 

1586: As before, the influence of those degrees of freedom whose energies lie in

1587: between $v_W/l$ and $v_C/d$ is shown by the appearance of the term 

1588: $(T_l/T_d)^{2({\tilde K}_{eff}-1)}$. This is because these degrees 

1589: of freedom have also been integrated away during the RG procedure, and 

1590: there must be continuity in the conductance expressions at $T=T_l$ whether we 

1591: come from the $T_d \gg T \gg T_l$ regime or the $T_l \ll T$ regime. Thus,

1592: we obtain the conductance in this regime as

1593: \beq

1594: G= {e^2\over h} [1 - {\tilde c}_3V_1^2 (\frac{T_d}{\Lambda})^{2(K_{eff}-1)}) 

1595: (\frac{T_l}{T_d})^{2({\tilde K}_{eff}-1)} (\frac{T}{T_l})^{2(K_L-1)}] 

1596: \quad {\rm for} \quad T\gg T_L ~,

1597: \eeq

1598: where ${\tilde c}_3$ is a constant similar in nature to ${\tilde c}_2$. We can 

1599: now see that, as $K_L=1$ (for 2DEG Fermi reservoirs), the conductance has no

1600: temperature dependence in the low temperature regime. 

1601: 

1602: A similar analysis can be done for the QPC limit, which reproduces the 

1603: conductance expressions for the QPC limit that were obtained explicitly in 

1604: the earlier subsection. We note, however, that the conductance 

1605: corrections are small in this case as the RG flow for the barriers 

1606: is restricted by the small length scales in the system.

1607: 

1608: The conductance corrections for electrons with spin can also be

1609: obtained using the effective actions and the RG equations for the 

1610: barriers, by proceeding in the same way as was done for spinless 

1611: electrons. Since the conductance expressions have already been

1612: given in the previous section, we do not repeat them here.

1613: 

1614: Thus, we emphasize that just by using the effective action and the RG 

1615: equations for the barriers, we can actually obtain the conductance 

1616: corrections. However, all that we actually do here is

1617: to compute the RG flows of the individual barriers, and then infer

1618: the temperature and length power-laws in the

1619: conductance corrections. Hence, even in principle, there is no way of 

1620: obtaining the constants ${\tilde c}_1 ,\ldots ,{\tilde c}_6$ from this method,

1621: whereas the explicit computation of the conductance in the earlier subsection

1622: can give the explicit forms of the constants as well. 

1623: In fact, the correlation functions computed there

1624: can be directly related to the coefficients which appear in the RG

1625: equations. In the various frequency regimes, the RG equation for a 

1626: single barrier for spinless fermions can be written as 

1627: \beq

1628: {dV\over d\lambda} = (1-2 |\bo| G_\bo (x,x))V. 

1629: \eeq 

1630: On Fourier transforming, this gives

1631: \beq

1632: \frac{dV (x)}{d\lambda} = ( 1 + 2\frac{dG}{d\lambda}(x,x,\tau_0 e^\lambda))

1633: V(x) ~,

1634: \eeq

1635: where $\tau_0$ is the high energy cutoff $1/\Lambda$. Integrating this 

1636: gives the renormalized strength of the impurity $V_{ren}$ as

1637: \beq

1638: V_{ren}(x,\lambda) = V(x,\lambda=0) \exp [\lambda - U(x,x,\tau_0 

1639: e^\lambda)] ~,

1640: \eeq

1641: where $U(x,x,\tau_0e^\lambda) = -2 (G_0(x,x,\tau_0 e^\lambda) - 

1642: G_0(x,x,\tau_0))$.

1643: Thus, in this case, 

1644: \beq

1645: \chi_1 (x,x) = \exp [l -U(x,x,\tau_0 e^\lambda )] ~.

1646: \eeq

1647: However, the non-local $\chi_1 (x,y)$ is not so easy to obtain just

1648: from the RG equations.

1649: 

1650: Now, let us study the conclusions that can be drawn from the conductance

1651: expressions. To begin with, the expressions in the various frequency regimes

1652: reveal that as {\it either} the temperature $T$ 

1653: is raised {\it or} the total length $L$ of the contacts and QW is decreased, 

1654: the conductance corrections become smaller and the conductance approaches

1655: integer multiples of $2g_0$ as expected \cite{tarucha,yacoby}. Furthermore, 

1656: we can see from these expressions that in the high temperature limit

1657: i.e., when $T \gg T_d, ~T_l$, the conductance corrections are independent 

1658: of the QW parameters. Hence, they are {\it independent} of the gate voltage 

1659: $V_G$ and of all factors dependent on the channel index. Thus they yield

1660: renormalizations to the ideal values which are themselves plateau-like

1661: and uniform for all channels. Such corrections to

1662: the conductance explain some of the puzzling features observed in the

1663: experiments of Ref. \cite{yacoby}. A more detailed 

1664: comparison of these results against experimental findings will be made 

1665: in a later section; it is important to note here that our results are in 

1666: qualitative agreement with most experimental observations on electronic 

1667: transport through a variety of quantum wire systems.

1668: 

1669: Let us now compare these observations with what we find as the 

1670: perturbative renormalizations to the perfect conductance of a barrier/impurity 

1671: placed anywhere within the quantum wire itself such that its distance from 

1672: the left lead-contact junction (taken as the origin) is again denoted by $a$. 

1673: An exactly similar computation of the one-point Green's function in this 

1674: case reveals that (note that we are now working with the hierarchy of 

1675: $d \ll a \ll l$)

1676: \beq

1677: G_{\bar{\omega}} (x=a,y=a) \simeq \frac{K}{2\vert\bar{\omega}\vert} ~,

1678: \eeq

1679: where

1680: \begin{displaymath}

1681: K = \left\{ \begin{array}{ll}

1682: K_W & \textrm{for $\vert\bar{\omega}\vert \gg v_C/d$}\\ \\

1683: K_W & \textrm{for $v_W/a \ll \vert\bar{\omega}\vert 

1684: \ll v_C/d$}\\ \\

1685: \frac{2K_LK_W}{K_L+K_W} & \textrm{for $v_W/l \ll \vert\bar{\omega}\vert 

1686: \ll v_C/a$}\\ \\

1687: K_L & \textrm{for $\vert\bar{\omega}\vert \ll v_W/l$ ~.} 

1688: \end{array} \right.

1689: \end{displaymath}

1690: Now, for a barrier at the left contact-QW junction, $a \rightarrow d$,

1691:  the second frequency regime of $v_W/a \ll \vert\bar{\omega}\vert 

1692: \ll v_C/d$ does not exist. We also

1693: find that $K = K'_{eff} \equiv 2K_CK_W/(K_C+K_W)$ rather than $K_W$ for

1694:  $\vert\bar{\omega}\vert \gg v_C/d$.

1695: Thus for a quantum wire system which has the two 

1696: contact regions and only barriers at the two contact-QW junctions, the 

1697: conductance in the highest temperature regime of $T \gg T_d$ is

1698: \beq

1699: g = g_0 K_L [ 1 - {\tilde c}_1 ({T \over \Lambda})^{2(K'_{eff} - 1)} 

1700: (|V(d)|^2 + |V(l+d)|^2) ] ~.

1701: \eeq

1702: Here ${\tilde c}_1$ is a dimensionful 

1703: constant which will depend on factors like $v_W$ and hence also the gate 

1704: voltage $V_G$. For the intermediate temperature regime of $T_l \ll T \ll T_d$, 

1705: we find the conductance to be

1706: \beq 

1707: g = g_0 K_L [ 1 - {\tilde c}_2 ({T_d \over \Lambda})^{2(K'_{eff} - 1)}

1708: ({T \over T_d})^{2({\tilde K}_{eff}-1)} (|V(d)|^2 + |V(l+d)|^2) ] ~,

1709: \eeq

1710: where ${\tilde K}_{eff} = 2K_L K_W/(K_L + K_W)$ as before, and ${\tilde c}_2$

1711: is also a dimensionful constant dependent on $V_G$. Finally, for the lowest 

1712: temperature regime of $T \ll T_l$, we obtain

1713: \beq

1714: g = g_0 K_L [ 1 - {\tilde c}_3 ({T \over T_l})^{2(K_L - 1)} ({T_d \over 

1715: \Lambda})^{2(K'_{eff} - 1)} ({T_l \over T_d})^{2({\tilde K}_{eff} - 1)}

1716: |V(0) + V(l+2d)|^2 ] ~,

1717: \eeq

1718: where ${\tilde c}_3$ too is a dimensionful constant dependent on $V_G$. Thus, 

1719: we can see that any barrier or impurity placed anywhere inside the QW will 

1720: always give a perturbative renormalization to the conductance which will be 

1721: dependent on the gate voltage and hence can {\it never} be flat or even 

1722: channel independent. The conductance expressions for the case of spinful 

1723: electrons can be found for this case in exactly the same way as before.

1724: 

1725: \section{\bf Effects of a Magnetic Field}

1726: 

1727: In this section, we will study the effects of an in-plane magnetic field on 

1728: the conductivity of a quantum wire. In general, a magnetic field couples to 

1729: both the spin (Zeeman coupling) and the orbital motion of an electron. 

1730: However, orbital motion is not possible in an in-plane magnetic field

1731: because the electrons are constrained to move only in the plane.

1732: Thus we will only consider the 

1733: effect of the Zeeman term. This term couples differently to spin up 

1734: and spin down electrons; here up and down are defined with respect to the 

1735: direction of the magnetic field which may or not be parallel to the quantum 

1736: wire. Thus the $SU(2)$ symmetry of rotations is explicitly broken. We will now 

1737: see that the spin and charge degrees of freedom do not decouple any longer.

1738: Our findings reveal that 

1739: 

1740: \noindent (a) for low magnetic fields (of about $0-3T$ for 

1741: Ga-As systems), the Zeeman splitting of the Fermi energies of the 

1742: two spin species of electrons in the QW is very small, and its effects 

1743: can be ignored. 

1744: 

1745: \noindent (b) for intermediate magnetic fields (of about $3-8T$),

1746: the Zeeman splitting of the Fermi energies of up and down spins becomes 

1747: appreciable; up and down spins see the two barrier

1748: strengths renormalize differently because of the Zeeman splitting, giving

1749: rise to an odd-even effect in the conductance of the two spin species.

1750: 

1751: \noindent (c) at still higher magnetic fields (of about $8-16T$),

1752: when each of the earlier sub-bands is completely Zeeman split into two

1753: spin-split sub-bands, conductance steps will be seen in multiples of 

1754: $g_0=e^2/h$; the odd-even effect will be most pronounced here with

1755: odd numbered spin-split sub-bands (containing only aligned

1756: moments) having a much less renormalized conductance and

1757: even numbered spin-split sub-bands (containing only

1758: anti-aligned moments) having a much more renormalized conductance, and 

1759: we can treat each spin-split sub-band as an effectively spinless TLL system.

1760: 

1761: \noindent (d) at magnetic fields much higher than this, all the spins in the 

1762: system will be spin polarized.

1763:  

1764: \subsection{The infinite TLL Quantum Wire and the Odd-Even Effect}

1765: 

1766: Let us first consider an infinitely long wire containing noninteracting 

1767: electrons. A magnetic field $h$ contributes the following term to the 

1768: Hamiltonian

1769: \beq

1770: - ~g \mu_B h ~S_{z,total} ~=~ - ~\frac{g \mu_B h}{2} ~(~ \rho_{0\ua} ~

1771: +~ \frac{1}{\sqrt \pi} ~\partial_x \phi_\ua ~-~ \rho_{0\da} ~-~ 

1772: \frac{1}{\sqrt \pi} ~\partial_x \phi_\da ~)~, 

1773: \label{magham}

1774: \eeq

1775: where $g$ is the gyromagnetic ratio (which is $2$ for free electrons but

1776: may be substantially smaller in quantum wire systems), $\rho_{0\ua} ,

1777: \rho_{0\da}$ respectively denote the mean density of the spin up 

1778: and spin down electrons, and $\phi_\ua , \phi_\da$ denote the 

1779: bosonic fields for the spin up and down electrons. The density terms

1780: $\rho_{0\sigma} = \rho_{0\ua} - \rho_{0\da}$ 

1781: have a bigger effect than the derivative terms $\partial_x

1782: \phi_\sigma$; by altering the chemical potentials for spin up and down

1783: electrons, these terms lead to different Fermi momenta 

1784: and therefore to different Fermi velocities $v_{F\ua}$

1785: and $v_{F\ua}$ for the two kinds of electrons. 

1786: 

1787: We now add a density-density interaction of the form ${\cal U} \rho^2 /2$ 

1788: where $\rho = \rho_\ua + \rho_\da$. (For instance, this may describe a 

1789: short-range Coulomb repulsion as in the Hubbard model; in that case $\cal U$ 

1790: is positive). The bosonized Lagrangian density takes the form

1791: \bea

1792: {\cal L} ~=~ & & \frac{1}{2v_{F\ua}} ~(\partial_t \phi_\ua )^2 ~-~

1793: \frac{v_{F\ua}}{2} ~(\partial_x \phi_\ua )^2 ~+~ 

1794: \frac{1}{2v_{F\da}} ~(\partial_t \phi_\da)^2 ~-~

1795: \frac{v_{F\da}}{2} ~(\partial_t \phi_\da)^2 \nonum \\

1796: & & -~ \frac{\cal U}{2\pi} ~(~\partial_x \phi_\ua ~+~ \partial_x 

1797: \phi_\da ~)^2 ~,

1798: \label{hamlag1}

1799: \eea

1800: where we have dropped some additive constants, and have only kept terms

1801: which are quadratic in the fields. We can rediagonalize this Lagrangian by 

1802: defining two new fields $\phi_i$, velocities $v_i$ and interaction parameters 

1803: $K_i$ (where $i=+,-$), and a mixing angle $\gamma$, where 

1804: \bea

1805: \phi_+ &=& {\sqrt {K_+ v_+}} ~(\frac{1}{\sqrt {v_{F\ua}}} \cos 

1806: \gamma ~\phi_\ua + \frac{1}{\sqrt {v_{F\da}}} \sin \gamma ~\phi_\da )

1807: \equiv p \phi_\ua + q \phi_\da , \nonum \\

1808: \phi_- &=& {\sqrt {K_- v_-}} ~( - \frac{1}{\sqrt {v_{F\ua}}} \sin 

1809: \gamma ~\phi_\ua + \frac{1}{\sqrt {v_{F\da}}} \cos \gamma ~\phi_\da )

1810: \equiv r \phi_\ua + s \phi_\da ,

1811: \label{transf1}

1812: \eea

1813: and

1814: \bea

1815: v_+^2 ~+~ v_-^2 ~&=&~ v_{F\ua}^2 ~+~ v_{F\da}^2 ~+~ \frac{\cal U}{\pi} ~

1816: (~ v_{F\ua} ~+~ v_{F\da} ~)~, \nonum \\

1817: (~ v_+^2 ~-~ v_-^2 ~)~ \cos ~(2\gamma) ~&=&~ v_{F\ua}^2 ~-~ 

1818: v_{F\da}^2 ~+~ \frac{\cal U}{\pi} ~ (~ v_{F\ua} ~-~ v_{F\da} ~)~, \nonum \\

1819: (~ v_+^2 ~-~ v_-^2 ~)~ \sin ~(2\gamma) ~&=&~ \frac{2{\cal U}}{\pi} ~{\sqrt {

1820: v_{F\ua} v_{F\da}}}~, \nonum \\

1821: K_+ v_+ ~(~ \frac{\cos^2 \gamma}{v_{F\ua}} ~+~ \frac{\sin^2 

1822: \gamma}{v_{F\da}} ~) ~&=& ~1 ~, \nonum \\

1823: K_- v_- ~(~ \frac{\sin^2 \gamma}{v_{F\ua}} ~+~ \frac{\cos^2 

1824: \gamma}{v_{F\da}} ~) ~&=& ~1 ~.

1825: \label{transf2}

1826: \eea

1827: The Lagrangian density in Eq. (\ref{hamlag1}) then takes the decoupled form

1828: \beq

1829: {\cal L} ~=~ \frac{1}{2K_+ v_+} ~(\partial_t \phi_+ )^2 ~-~ \frac{v_+}{2K_+} ~

1830: (\partial_x \phi_+ )^2 ~+~ \frac{1}{2K_- v_-} ~(\partial_t \phi_- )^2 ~-~

1831: \frac{v_-}{2K_-} ~(\partial_t \phi_- )^2 ~.

1832: \label{hamlag2}

1833: \eeq

1834: Thus the charge and spin degrees of freedom get mixed since the fields 

1835: $\phi_+$ and $\phi_-$ which diagonalize the Lagrangian will generally be 

1836: different from the fields $\phi_\rho = (\phi_\ua + \phi_\da)/\sqrt{2}$

1837: and $\phi_\sigma = (\phi_\ua - \phi_\da)/\sqrt{2}$. Note that if the 

1838: magnetic field $h$

1839: is zero, then $v_{F\ua} = v_{F\da}$ and $\gamma = \pi /4$; $\phi_+$ and

1840: $\phi_-$ are then identical (up to a sign) to the charge and spin fields 

1841: $\phi_\rho$ and $\phi_\sigma$.

1842: 

1843: We will now present the RG equations for a weak $\delta$-function 

1844: impurity placed at the origin. For this, we need to 

1845: compute the scaling dimension of the impurity term 

1846: in the Lagrangian. We first write the impurity term in terms of fermionic

1847: fields $\psi_{\ua}(0)$ and $\psi_{\da}(0)$, and then in terms of 

1848: the bosonic fields $\phi_{\ua}(0)$ and $\phi_{\da}(0)$ at the origin as

1849: \bea

1850: L_{imp} &=& V(0) (\psi_{\ua}^{\dag}(0) \psi_{\da}(0)

1851: + \psi_{\da}^{\dag}(0) \psi_{\ua}(0)) \nonum \\

1852: &=& V_1 \Lambda ~[\cos(2\sqpi\phi_{\ua}(0)) +\cos(2\sqpi\phi_{\da}(0))] ~.

1853: \eea

1854: We then invert the relations between $\phi_{\pm}$ and 

1855: $\phi_{\ua,\da}$ given above in order to rewrite the above 

1856: expression for the impurity in terms of the diagonal fields $\phi_{\pm}$. 

1857: The scaling dimensions of the impurity terms are then found to be

1858: \beq

1859: D_{V_{\ua}} = v_{F\ua} (\frac{\cos^{2}\gamma}{v_+} 

1860: + \frac{\sin^{2}\gamma}{v_-}) ~~~,~~~ D_{V_{\da}} = 

1861: v_{F\da} (\frac{\sin^{2}\gamma}{v_+} + \frac{\cos^{2}\gamma}{v_-}).

1862: \eeq

1863: Hence the RG equations for $V_{\ua}$ and $V_{\da}$ are given by

1864: \bea

1865: \frac{dV_{\ua}}{d\lambda} &=& ( 1 - D_{V_{\ua}}) V_{\ua} \nonum \\

1866: \frac{dV_{\da}}{d\lambda} &=& ( 1 - D_{V_{\da}}) V_{\da} ~,

1867: \eea

1868: where $V_\ua$ and $V_\da$ both start from the value $V_1$ at the microscopic

1869: length scale. 

1870: 

1871: We can now study what happens in the presence of strong and 

1872: weak magnetic fields. But let us first remind ourselves of the following 

1873: relations (which result from the Zeeman splitting),

1874: \bea

1875: v_{F\ua,\da} &=& v_F \sqrt{1 \pm \frac{g\mu_{B} h}{2E_{F1D}}} \nonum \\

1876: \tan(2\gamma) &=& \frac{2{\cal U} \sqrt{v_{F\ua}v_{F\da}}/\pi}

1877: {(v_{F\ua}-v_{F\da})(v_{F\ua}+v_{F\da} + {\cal U}/\pi)} ~,

1878: \eea 

1879: where $v_F ={\sqrt {2E_{F1D}/m}}$ is the Fermi velocity in the absence of a 

1880: magnetic field. Therefore, in the limit of a strong magnetic field where 

1881: the Zeeman splitting of the two spin species is much larger 

1882: than the short ranged interaction energy $\cal U$ (i.e., ${\cal U} \ll \vert 

1883: v_{F\ua}-v_{F\da}\vert$ and $\gamma \ll \pi/2$), we can approximate 

1884: the relations for the two velocities 

1885: $v_{\pm}$ (to linear order in ${\cal U}/(v_{F\ua}-v_{F\da})$) as

1886: \bea 

1887: v_+ &=& v_{F\ua} + \delta_+ ~~~,~~~ \delta_+\ll v_{F\ua} \nonum \\

1888: v_- &=& v_{F\da} + \delta_- ~~~,~~~ \delta_-\ll v_{F\da} ~.

1889: \eea

1890: By putting these relations for $v_{\pm}$ into the expressions given 

1891: above relating $v_{\pm}$ and $v_{F\ua,\da}$, we get 

1892: \beq 

1893: \delta_+ ~=~ \delta_- ~=~ \frac{\cal U}{2\pi} ~.

1894: \eeq

1895: Now, using this (together with the fact that $\gamma \ll \frac{\pi}{2}$) in 

1896: the two RG equations obtained above gives us

1897: \bea

1898: \frac{dV_{\ua}}{d\lambda} &=& (1 - \frac{v_{F\ua}}{v_+}) V_{\ua} \nonum \\

1899: &\simeq& \frac{\cal U}{2\pi v_{F\ua}} V_{\ua} ~,

1900: \eea

1901: and

1902: \bea

1903: \frac{dV_{\da}}{d\lambda} &=& ( 1 - \frac{v_{F\da}}{v_-}) V_{\da} \nonum \\

1904: &\simeq& \frac{\cal U}{2\pi v_{F\da}} V_{\da} ~.

1905: \eea

1906: Now, these two RG equations indicate that as $v_{F\ua}$ is larger than 

1907: $v_{F\da}$, the renormalized impurity strength felt by those electrons 

1908: which have their magnetic moments aligned with the external B field is less 

1909: than the renormalized impurity strength felt by the electrons which have their 

1910: magnetic moments anti-aligned with the B field. Furthermore, if the B field 

1911: is further increased, we will reach a situation when alternate sub-bands 

1912: in the QW will be populated by either only the aligned or only the 

1913: anti-aligned electrons. In this regime, the difference in back-scattering 

1914: felt by the two species of electrons will be very clear from the alternating 

1915: weak and strong corrections to the conductance. To be more specific, all odd 

1916: numbered sub-bands will show much less corrections to the perfect conductance 

1917: (as they will be populated by electrons aligned with the magnetic field), 

1918: while all even numbered 

1919: sub-bands will show much greater corrections to the perfect conductance 

1920: (as they will be populated by the anti-aligned electrons). This 

1921: {\it odd-even} effect had, in fact, been predicted by a two-band TLL study of 

1922: Kimura {\it et al} \cite{kimura}, but its explanation 

1923: on the grounds of impurity 

1924: renormalization is now made clear. Furthermore, this effect has been recently 

1925: observed by Liang {\it et al} \cite{liang}, and we will discuss their 

1926: observations in a later section. It should be noted here that though the 

1927: odd-even effect is easy to show upon taking the limit of ${\cal U} \ll 

1928: \vert v_{F\ua} 

1929: - v_{F\da} \vert$, the existence of this phenomenon does not need 

1930: this limit to be taken. Furthermore, we also find that upon taking the limit 

1931: of $h \ll E_{F1D}$ (i.e., weak magnetic field with $v_{F\ua} = v_{F\da} =

1932: v_F$), 

1933: \beq

1934: \frac{dV_{\ua}}{d\lambda} = \frac{dV_{\da}}{d\lambda} = \frac{\cal U}{2\pi 

1935: v_F} ~,

1936: \eeq

1937: which tells us that the odd-even effect vanishes in the weak magnetic field 

1938: limit. Finally, we comment on the fact that the odd-even effect discussed 

1939: above gives rise to the possibility of the creation of a spin-valve (i.e., 

1940: a spin polarized current creating device) in these QW systems. Even though 

1941: the odd-even effect needs a high magnetic field to be observed in current 

1942: day experiments \cite{liang}, it may be possible to employ artificial barriers 

1943: like negatively-biased finger gates to heighten the difference in 

1944: renormalization of the up and down spin electrons at lower magnetic fields. 

1945: At this point, however, quantitative predictions are difficult to make.

1946:  

1947: \subsection{A study of our model for the QW with a magnetic field}

1948: 

1949: Having discussed how to obtain a diagonal Lagrangian when both a magnetic

1950: field and interactions are present as well as shown the interesting odd-even 

1951: effect that takes place because of an impurity in an infinite TLL in the 

1952: presence of an external B field, we will now study what happens when the 

1953: model in Sec. 2 is placed in a magnetic field. In the regions $x < 0$ and 

1954: $x > l+2d$, we have a system of noninteracting electrons parametrized by 

1955: velocities $v_{F\ua}$ and $v_{F\da}$. In the regions of the 

1956: contacts $0<x<d$ and $l+d<x<l+2d$, we have an interacting system parametrized 

1957: by two velocities $v_{C+}$, $v_{C-}$ and a mixing angle $\gamma_C$. In the

1958: quantum wire $d<x<l+d$, the system is parametrized by the velocities $v_{W+}$, 

1959: $v_{W-}$ and a mixing angle $\gamma_W$. The last six parameters are functions 

1960: of $v_{F\ua}$, $v_{F\da}$ and the strengths of the interaction in the contacts

1961: and quantum wire. The action for this model is given by

1962: \bea

1963: S_0 &=& \int d\tau ~[\int_{-\infty}^0 +\int_{l+2d}^\infty]~ dx ~\{ ~{1 \over 

1964: 2K_L}[ {1\over v_{F\ua}} (\pt\phi_\ua)^2 + v_{F\ua} (\px\phi_\ua)^2 ]~

1965: +~ [\ua \rightarrow \da ] ~\} \nonum \\

1966: &+& \int d\tau ~[\int_{0}^d +\int_{l+d}^{l+2d}] ~dx ~\{ ~{1 \over 2 K_{C+}}

1967: [{1\over v_{C+}} (\pt\phi_{C+})^2 + v_{C+} (\px \phi_{C+})^2 ] ~+~ 

1968: [ C+ \rightarrow C- ] ~\} \nonum \\

1969: &+& \int d\tau ~\int_{d}^{l+d} ~dx ~\{ ~{1\over 2 K_{W+} }[{1\over v_{W+}} 

1970: (\pt \phi_{W+})^2 + v_{W+} (\px\phi_{W+})^2 ] ~+~ [W+ \rightarrow W- ] ~\} ~. 

1971: \nonum \\

1972: &&

1973: \label{actmag}

1974: \eea

1975: Note that we have 

1976: ignored the junction barriers and the gate voltage for the moment. It is 

1977: worth mentioning here that we have verified, by performing a calculation of 

1978: the kind outlined in Ref. \cite{maslov}, that our model for the 

1979: QW system when placed in an external magnetic field and in the absence of

1980: any barriers/impurities still gives perfect conductance in the dc limit for 

1981: each sub-band. 

1982: 

1983: We begin by noting that we will present here the calculation for the 

1984: case when the mixing angle $\gamma$ is the same in both the contacts as well 

1985: as the QW, i.e., the short ranged electron-electron interaction $\cal U$ is 

1986: equal in all the three TLLs. Though this is not necessarily the case in a real 

1987: system, we will present it as it considerably simplifies the computations 

1988: while providing us with an adequate discussion of all the important results 

1989: for effective actions, conductance expressions, resonances, etc. 

1990: Later, we will briefly discuss the case in which

1991: the mixing angle is different in the contact and QW regions. The explicit 

1992: derivation of the effective action is given in Appendix C.

1993: The high frequency effective Lagrangian density ($\bo \gg v_{C\pm}/d , 

1994: v_{W\pm}/l$) simplifies to 

1995: \bea

1996: {\cal L}_{eff, high} &=& {|\bo| \over 2}[(\frac{1}{K_L} + 

1997: \frac{p^2}{K_{C+}} + \frac{r^2}{K_{C-}})(\tphi_{1\ua}^2 + \tphi_{4\ua}^2) 

1998: + (\frac{1}{K_L} + \frac{q^2}{K_{C+}} + \frac{s^2}{K_{C-}})

1999: (\tphi_{1\da}^2 + \tphi_{4\da}^2) \nonum \\

2000: &&~~~~~ +~2(\frac{pq}{K_{C+}}+\frac{rs}{K_{C-}})(\tphi_{1\ua}

2001: \tphi_{1\da} + \tphi_{4\ua}\tphi_{4\da})] \nonum \\

2002: && +~~{|\bo| \over 2}[\{p^2(\frac{1}{K_{C+}}+\frac{1}{K_{W+}}) +r^2(

2003: \frac{1}{K_{C-}}+\frac{1}{K_{W-}})\}(\tphi_{2\ua}^2 +\tphi_{3\ua}^2) \nonum \\

2004: &&~~~~~~~~~~ +~ \{q^2(\frac{1}{K_{C+}}+\frac{1}{K_{W+}}) +s^2(

2005: \frac{1}{K_{C-}}+\frac{1}{K_{W-}})\}(\tphi_{2\da}^2 +\tphi_{3\da}^2) \nonum \\

2006: &&~~~~~~~~~~ +~ 2\{pq(\frac{1}{K_{C+}}+\frac{1}{K_{W+}})

2007: +rs(\frac{1}{K_{C-}}+\frac{1}{K_{W-}})\}(\tphi_{2\ua}\tphi_{2\da} 

2008: + \tphi_{3\ua}\tphi_{3\da})] \nonum \\

2009: && +~~ {\cal L}_{gate}(\phi_{2\ua,\da},

2010: \phi_{3\ua,\da}) + {\cal L}_{imp}(\phi_{1\ua,\da}, \phi_{4\ua,\da}) .

2011: \eea 

2012: We can clearly see the separation between the outer two and inner two fields. 

2013: This means that we can integrate out the inner fields 

2014: $\tphi_{2\ua,\da}$ and $\tphi_{3\ua,\da}$ without any 

2015: further work and be left with a high frequency effective action dependent 

2016: on $\tphi_{1\ua,\da}$ and $\tphi_{4\ua,\da}$ exactly as 

2017: given above (and without any influence of the gate voltage $V_G$ either). 

2018: We can also make the prediction that the conductance corrections due to 

2019: barriers at the outer two junctions will have temperature power-laws which 

2020: will be combinations of $K_L$ and $K_{C\pm}$ (much like those seen before) and 

2021: also that it will be not be dependent on the gate voltage. 

2022: 

2023: In the intermediate frequency range of $v_{W\pm}/l \ll \bo \ll v_{C\pm}/d$, we 

2024: get (after integrating out the inner fields $\tphi_{2\ua,\da}$ and 

2025: $\tphi_{3\ua,\da}$), 

2026: \bea

2027: {\cal L}_{eff,int} &=& {|\bo| \over 2}[(\frac{1}{K_L} + \frac{p^2}{K_{W+}} 

2028: + \frac{r^2}{K_{W-}})(\tphi_{1\ua}^2 + \tphi_{4\ua}^2) 

2029: + (\frac{1}{K_L} + \frac{q^2}{K_{W+}} + \frac{s^2}{K_{W-}})(\tphi_{1\da}^2 

2030: + \tphi_{4\da}^2) \nonum \\

2031: &&~~~~ +~2(\frac{pq}{K_{W+}}+\frac{rs}{K_{W-}})(\tphi_{1\ua}

2032: \tphi_{1\da} + \tphi_{4\ua}\tphi_{4\da})] \nonum \\

2033: && + \frac{{\tilde V}_G (p-q)(s-r)}{(ps-qr)}(\phi_{4\ua} - \phi_{1\ua} +

2034: \phi_{4\da} - \phi_{1\da}) ~,

2035: \eea

2036: where $\alpha_{C\pm} = v_{C\pm}/(K_{C\pm}d)$ are the charging energies for 

2037: the $\tphi_{\pm}$ fields, and they appear because of the growth of the 

2038: coherence in the system over the contact regions. We can now predict that 

2039: the conductance corrections due to barriers at the outer two junctions will

2040: have temperature power-laws which will be combinations of $K_L$ and 

2041: $K_{W\pm}$ (again like those seen previously), and that this correction will 

2042: definitely be dependent on the gate voltage. 

2043: 

2044: Finally, in the low frequency 

2045: limit of $\bo \ll v_{W\pm}/l$, we get

2046: \bea

2047: {\cal L}_{eff,low} &=& {|\bo| \over 2K_L} (\tphi_{1\ua}^2 + \tphi_{1\da}^2 + 

2048: \tphi_{4\ua}^2 +\tphi_{4\da}^2) \nonum \\ 

2049: && +\sum_{\pm} {v_{C\pm}\over 2d}[\phi_{2\pm}^2 +\phi_{3\pm}^2 +

2050: \phi_{1\pm}^2 + \phi_{4\pm}^2 -2\phi_{2\pm} \phi_{1\pm} -2\phi_{3\pm} 

2051: \phi_{4\pm}] \nonum \\

2052: && +\sum_\pm {v_{W\pm}^2\over 2l v_{C\pm}}(\phi_{2\pm} -\phi_{3\pm})^2 

2053: \nonum \\

2054: && +\sum_\pm \frac{{\tilde V}_G}{2(ps-qr)}[(s-r)(\phi_{3+}-\phi_{2+}) + 

2055: (p-q)((\phi_{3-}-\phi_{2-})] .

2056: \eea

2057: Integrating out the inner fields, we are then left with an effective action in 

2058: terms of the new current and charge variables respectively 

2059: \bea

2060: \phi_{14\ua,\da} &=& \frac{\phi_{1\ua,\da} + \phi_{4\ua,\da}}{2} \nonum \\

2061: n_{14\ua,\da} &=& \frac{\phi_{4\ua,\da} - \phi_{1\ua,\da}}{\sqpi} ~.

2062: \eea

2063: that span the coherent TLL system

2064: between the two lead-contact junction barriers in this 

2065: low frequency regime. Thus, we get the effective Lagrangian density in this 

2066: regime as

2067: \bea

2068: {\cal L}_{eff,low} &=& {|\bo| \over K_L}[\tphi_{14\ua}^2 + \tphi_{14\da}^2 + 

2069: \frac{\pi}{4} {\tilde n}_{14\ua}^2 + \frac{\pi}{4} {\tilde n}_{14\da}^2] 

2070: \nonum \\

2071: && + \frac{U_1}{2}(p n_{14\ua} + q n_{14\da} + n_{01})^2 + \frac{U_2}{2}(r 

2072: n_{14\ua} + s n_{14\da} + n_{02})^2 + {\cal L}_{imp},

2073: \eea

2074: where 

2075: \bea U_1 &=& \frac{\pi v_{W+}\alpha_{+}\beta_{+}} {v_{C+}\alpha_+ 

2076: + 2v_{W+}\beta_+} ~~,~~ U_2 = \frac{\pi v_{W-}\alpha_{-}\beta_{-}}

2077: {v_{C-}\alpha_- + 2v_{W-}\beta_-} \nonum \\

2078: n_{01} &=& \frac{{\tilde V}_G v_{C+}(s-r)}{\pi v_{W+}\beta_+ (ps-qr)} ~~,~~ 

2079: n_{02} = \frac{{\tilde V}_G v_{C-}(p-q)}{\pi v_{W-}\beta_- (ps-qr)} \nonum \\

2080: {\rm and} \quad \alpha_{\pm} &=& \frac{v_{C\pm}}{K_{C\pm}d} ~~,~~ 

2081: \beta_{\pm} = \frac{v_{W\pm}}{K_{W\pm}d},

2082: \eea

2083: and the term coming from the two barriers can be written as

2084: \beq

2085: {\cal L}_{imp} = 2V \Bigl( \cos(2\sqpi\phi_{14\ua}) \cos(\pi n_{14\ua}) + 

2086: \cos(2 \sqpi\phi_{14\da}) \cos(\pi n_{14\da}) \Bigr) ~.

2087: \eeq

2088: It becomes clear from the above expression that the temperature power-law 

2089: will be dependent on the lead interaction parameter $K_L = 1$, and the 

2090: conductance correction will, in this regime, be temperature independent (as 

2091: seen previously). Furthermore, the length corrections will be gate 

2092: voltage dependent.

2093: Let us now study the possible resonance symmetries of the low frequency 

2094: effective action given above. Even though the structure of this expression 

2095: is more complicated than those encountered previously, we can rewrite the 

2096: two charging terms as follows:

2097: \beq 

2098: {\cal L}_{charging} = \frac{(U_1 p^2 + U_2 r^2)}{2} [n_{14\ua} 

2099: + a n_{14\da} -b]^2 + f(n_{14\da}) + {\rm const} ~,

2100: \eeq

2101: where 

2102: \bea

2103: a &=& \frac{U_1 pq + U_2 rs}{U_1 p^2 + U_2 r^2} \nonum \\

2104: b &=& \frac{U_1 n_{01} + U_2 n_{02}}{U_1 p^2 + U_2 r^2} ~,

2105: \eea

2106: and $f(n_{14\da})$ is a quadratic function of the field 

2107: $n_{14\da}$ only. Thus, we can see that if we set 

2108: \beq

2109: a n_{14\da} - b = - Z - \frac{1}{2} ~,

2110: \eeq

2111: we get a resonance in the $n_{14\ua}$ parts of the charging and barrier 

2112: terms whenever $n_{14\ua} = Z$ or $Z+1$ and one makes the transformation 

2113: of $\phi_{14\ua}\rightarrow \phi_{14\ua}\pm \sqpi/2$. This means that 

2114: only the transport of all up-spin electrons through the two barriers 

2115: is at resonance, and this is clearly a one-parameter tuned resonance. 

2116: A one-parameter tuned resonance for only the transport of all down-spin 

2117: electrons through the two barriers can be found in exactly the same manner 

2118: by rewriting the above charging expressions but for $n_{14\da}$ 

2119: instead of $n_{14\ua}$. We also find another resonance given by

2120: \bea

2121: \frac{q}{p}n_{14\da} + \frac{1}{p}n_{01} &=& - Z - \frac{1}{2} ~, \nonum \\ 

2122: {\rm and} \quad \frac{s}{r}n_{14\da} + \frac{1}{r}n_{02} &=& - Z - 

2123: \frac{1}{2} ~,

2124: \eea 

2125: where $Z$ is the same integer in both equations; then there exists a possible 

2126: resonance whenever $n_{14\ua} = Z$ or $Z+1$, and one makes the 

2127: transformation of $\phi_{14\ua,\da}\rightarrow

2128: \phi_{14\ua,\da}\pm \sqpi/2$. This resonance symmetry would 

2129: lead to a vanishing of the barrier terms 

2130: in the effective action, and would 

2131: correspond to the transfer of an electron across the system. One 

2132: can immediately see that the above two conditions on $n_{14\da}$ 

2133: mean that two parameters, here the gate voltage $V_G$ and the external 

2134: magnetic field $h$, have to be manipulated to achieve this resonance 

2135: condition. Such a resonance will, therefore, be much more difficult to 

2136: observe experimentally. However, this resonance will lead to a complete 

2137: vanishing of all back-scattering (and hence the conductance corrections) 

2138: while the two one-parameter tuned resonances will give only partial 

2139: lessening of the conductance corrections. 

2140: 

2141: The two one-parameter tuned resonances can prove useful in 

2142: creating a spin-valve. Even at a low magnetic field, the odd-even effect 

2143: can be enhanced by using stronger artificial barriers (e.g., by

2144: employing finger gates over the channel) or making the length of the channel 

2145: longer or working at lower temperatures, together with tuning the

2146: transport of only one spin species of electrons through the two barriers at 

2147: resonance. Thus one can create an enhanced spin polarized electron current 

2148: output from the QW system.

2149: 

2150: Before we go on to computing the conductance through the system for the above 

2151: model in the presence of the magnetic field, let us make some remarks about 

2152: the case when the mixing angle in the QW is taken to be different from that 

2153: in the contacts. A long calculation does give expressions for the 

2154: effective action in the three frequency regimes similar to those obtained 

2155: above, but with two sets of the transformation coefficients 

2156: relating the $\phi_{\pm}$ and $\phi_{\ua,\da}$ fields. However, 

2157: the integrating out of the inner fields is a far more difficult task; 

2158: furthermore, the analysis reveals that the only possible resonance symmetry 

2159: of the low frequency effective action is one that needs at least four 

2160: parameters to be manipulated. We will, therefore, not present these 

2161: results as we do not find anything substantially 

2162: new from the analysis compared to the simpler case of equal mixing angles.

2163: 

2164: \subsection{Conductance of our model for the QW with a magnetic field}

2165: 

2166: We will begin by a re-writing the RG equation, obtained by Safi and Schulz 

2167: \cite{safilong} for an impurity placed within a QW of a finite size and 

2168: connected to Fermi leads, in a way which will be convenient to use 

2169: in computing the conductance expressions for our model of the QW with 

2170: barriers even in the presence of the external magnetic field. We begin by 

2171: quoting the expression for the RG flow found for an impurity in Ref.

2172: \cite{safilong},

2173: \beq

2174: \frac{dV_{m_{\rho},m_{\sigma}}(x)}{d\lambda} = [1 - \frac{1}{2}(m_{\rho}^2 

2175: \frac{dU_{\rho} (x,x,\tau_0 e^\lambda)}{d\lambda} + m_{\sigma}^2 \frac{dU_{

2176: \sigma} (x,x,\tau_0 e^\lambda)}{d\lambda})] V_{m_{\rho},m_{\sigma}}(x),

2177: \eeq

2178: where

2179: \bea

2180: U_{\rho}(x,x,\tau_0 e^\lambda) &=& 2 [G_{\rho}(x,x,\tau_0) - G_{\rho}

2181: (x,x,\tau_0 e^\lambda)] \nonum \\

2182: U_{\sigma}(x,x,\tau_0 e^\lambda) &=& 2 [G_{\sigma}(x,x,\tau_0) 

2183: - G_{\sigma}(x,x,\tau_0 e^\lambda)].

2184: \eea

2185: Now with $\phi_{\rho} = (\phi_{\ua} + \phi_{\da})/{\sqrt 2}$ and 

2186: $\phi_{\sigma} = (\phi_{\ua} - \phi_{\da})/{\sqrt 2}$, we can write

2187: \bea

2188: G_{\rho}(x,x,\tau_0 e^\lambda) &=& <\phi_{\rho}(x,\tau_0 e^\lambda)

2189: \phi_{\rho}(x,0)> = \frac{1}{2}[<\phi_{\ua}\phi_{\ua}> + <\phi_{\da}

2190: \phi_{\da}> + 2<\phi_{\ua}\phi_{\da}>] \nonum \\

2191: G_{\sigma}(x,x,\tau_0 e^\lambda) &=& <\phi_{\sigma}(x,\tau_0 e^\lambda)

2192: \phi_{\sigma}(x,0)> = \frac{1}{2}[<\phi_{\ua}\phi_{\ua}> + <\phi_{\da}

2193: \phi_{\da}> - 2<\phi_{\ua}\phi_{\da}>], \nonum \\

2194: &&

2195: \eea

2196: where the space-time indices are implicit on the right hand sides. 

2197: Substituting the expressions 

2198: for $U_{\rho}$ and $U_{\sigma}$ given above in the RG equation, and 

2199: working with the case for the back-scattering of one electron 

2200: $m_{\rho} = m_{\sigma} = 1$, we write the RG equation as

2201: \beq

2202: \frac{dV_{1,1}(x)}{d\lambda} = [1 + \frac{d}{d\lambda}(G_{\ua}(x,x,\tau_0 

2203: e^\lambda) + G_{\da}(x,x,\tau_0 e^\lambda))] V_{1,1}(x) ~,

2204: \eeq

2205: where $G_{\ua} = <\phi_{\ua}\phi_{\ua}>$ and 

2206: $G_{\da} = <\phi_{\da}\phi_{\da}>$. We will now

2207: use the effective actions found in the various frequency regimes 

2208: to obtain the two Green's functions $G_{\ua}$ and $G_{\da}$, 

2209: put these in the RG equations and thereby infer the corrections 

2210: to the conductance caused by the junction barriers. 

2211: 

2212: We start with the high and intermediate frequency/temperature effective 

2213: actions given earlier for the model when the mixing angle is the same in the 

2214: contact and QW regions. Here, we can 

2215: see that the final effective action (in terms of only the fields at the outer 

2216: two junctions) is the sum of two distinct parts, each of which is an 

2217: expression of the kind $\frac{A}{2}\phi_{\ua}^2 + 

2218: \frac{B}{2}\phi_{\da}^2 + C\phi_{\ua}\phi_{\da}$ 

2219: separately for fields $\phi_1$ and $\phi_4$. This tells us that we can simply 

2220: take the sum of the contributions from each of the two incoherent barriers. 

2221: Thus, the general expression 

2222: \beq

2223: {\cal L}_{eff} = \frac{A}{2\vert\bo\vert}\tphi_{\ua}^2 + \frac{B}{2\vert\bo

2224: \vert}\tphi_{\da}^2 + \frac{C}{\vert\bo\vert}\tphi_{\ua}\tphi_{\da},

2225: \eeq

2226: can be diagonalized in terms of two new fields $\phi_a$ and 

2227: $\phi_b$ i.e., written as

2228: \beq

2229: {\cal L}_{eff} = \frac{1}{2\vert\bo\vert} (\lambda_a {\tphi_a}^2 

2230: + \lambda_b {\tphi_b}^2) ~,

2231: \eeq

2232: where $\lambda_a$ and $\lambda_b$ are the eigenvalues of the transformation

2233: given by

2234: \beq

2235: \lambda_{a,b} = \frac{A+B}{2} \pm \frac{1}{2}\sqrt{(A-B)^2+C^2} ~.

2236: \eeq 

2237: Then, 

2238: \bea

2239: G_{\bo a} &=& <\tphi_a \tphi_a> = \frac{1}{\lambda_a 

2240: \vert\bo\vert} \nonum \\

2241: G_{\bo b} &=& <\tphi_b \tphi_b> = \frac{1}{\lambda_b \vert\bo\vert}.

2242: \eea

2243: Using the two eigenvectors corresponding to these two eigenvalues, we obtain 

2244: $G_{\bo\ua}$ and $G_{\bo\da}$ as

2245: \beq

2246: G_{\bo\ua} = \frac{C^2}{\vert\bo\vert \{(\lambda_a-A)(\lambda_b-B)-C^2\}^2} 

2247: [C^2 G_{\bo a} + (\lambda_b - B)^2 G_{\bo b}] ,

2248: \eeq

2249: and

2250: \beq

2251: G_{\bo\da} = \frac{C^2}{\vert\bo\vert \{(\lambda_a-A) (\lambda_b-B)-C^2\}^2} 

2252: [(\lambda_a-A)^2 G_{\bo a} + C^2 G_{\bo b}].

2253: \eeq

2254: We finally obtain an expression for $G_{\bo\ua} + G_{\bo\da}$ as

2255: \bea

2256: G_{\bo\ua} + G_{\bo\da} &=& \frac{C^2}{\vert\bo\vert \{(\lambda_a-A)

2257: (\lambda_b-B)-C^2\}^2} [\frac{C^2 + (\lambda_a - A)^2}{\lambda_a}

2258: + \frac{C^2 + (\lambda_b - B)^2}{\lambda_b}] \nonum \\

2259: &\equiv& \frac{K_{mag}}{2\vert\bo\vert} .

2260: \eea

2261: We can now use the Fourier transform of the above expression to obtain 

2262: the temperature and length power-laws for the conductance corrections in 

2263: the high and 

2264: intermediate frequency regimes. In the high temperature regime of 

2265: $T \gg T_d ~(\sim v_{C\pm}/d)$, we get the conductance as 

2266: \beq

2267: g = g_0 K_L [1 - c_1(|V(0)|^2 + |V(l+2d)|^2)(\frac{T}{\Lambda})^{(K_{eff,mag} 

2268: - 2)}] ,

2269: \eeq

2270: where $c_1$ is a dimensionful constant independent of the gate voltage $V_G$,

2271: and $K_{eff,mag}$ is given by the expression for $K_{mag}$ where 

2272: the coefficients $A$, $B$ and $C$ are given by

2273: \bea

2274: A &=& \frac{1}{K_L} + \frac{p^2}{K_{C+}} + \frac{r^2}{K_{C-}} \nonum \\

2275: B &=& \frac{1}{K_L} + \frac{q^2}{K_{C+}} + \frac{s^2}{K_{C-}} \nonum \\

2276: C &=& \frac{pq}{K_{C+}} + \frac{rs}{K_{C-}} ~.

2277: \eea

2278: Similarly, the conductance expression for the intermediate temperature defined 

2279: by $T_l ~(\sim v_{W\pm}/l) \ll T \ll T_d$ is given by 

2280: \beq

2281: g = g_0 K_L [1 - c_2(|V(0)|^2 + |V(l+2d)|^2)(\frac{{\tilde T}_d}{\Lambda})^{

2282: (K_{eff,mag} -2)}(\frac{T}{{\tilde T}_d}) ^{({\tilde K}_{eff,mag} - 2)}],

2283: \eeq

2284: where $c_2$ is another dimensionful constant which is dependent on the gate 

2285: voltage, $T_d$ has replaced $\Lambda$ as the correct cutoff for the 

2286: temperature, and ${\tilde K}_{eff,mag}$ is found in exactly the same way as 

2287: $K_{eff,mag}$ but with coefficients $A_1$, $B_1$ and $C_1$ defined as

2288: \bea

2289: A_1 &=& \frac{1}{K_L} + \frac{p^2}{K_{W+}} + \frac{r^2}{K_{W-}} \nonum \\

2290: B_1 &=& \frac{1}{K_L} + \frac{q^2}{K_{W+}} + \frac{s^2}{K_{W-}} \nonum \\

2291: C_1 &=& \frac{pq}{K_{W+}} + \frac{rs}{K_{W-}}.

2292: \eea

2293: Finally, we obtain the low frequency conductance expression for the 

2294: temperature regime of $T \ll T_l$ as 

2295: \beq

2296: g = g_0 K_L [1 - c_3(|V(0) + V(l+2d)|^2)(\frac{{\tilde T}_d}{\Lambda})^{

2297: (K_{eff,mag} -2)}(\frac{T_l}{T_d})^{({\tilde K}_{eff,mag} - 2)}

2298: (\frac{T}{T_l})^{2(K_L - 1)}],

2299: \eeq

2300: where $c_3$ is a dimensionful constant similar in nature to $c_2$, i.e., 

2301: dependent on gate voltage. This expression is also independent of the 

2302: temperature for Fermi leads with $K_L = 1$, and the coherence between the 

2303: barriers means that this correction term could go to zero at resonance. 

2304: 

2305: We end by noting that we can again take the limit of ${\cal U} \ll 

2306: \vert v_{C\ua} - v_{C\da} \vert$ in our equations to highlight the 

2307: existence of the odd-even effect within our model of the QW as well. Upon 

2308: taking this limit in the high temperature regime, we find that

2309: \beq

2310: \frac{dV_{\ua}}{d\lambda} \simeq \frac{\cal U}{4\pi v_{C\ua}},

2311: \eeq

2312: while

2313: \beq

2314: \frac{dV_{\da}}{d\lambda} \simeq \frac{\cal U}{4\pi v_{C\da}},

2315: \eeq 

2316: where $\cal U$ is the inter-electron interaction term. This clearly shows 

2317: that as $v_{C\ua}$ increases and $v_{C\da}$ 

2318: decreases with an increasing magnetic field, the renormalized barrier seen 

2319: by the two spin species of electrons will be different. We also note that, 

2320: just like the case of the infinite, homogeneous QW, a weak field of $h \ll 

2321: E_{F1D}$ in our model of the QW does not give rise to the odd-even effect. 

2322: 

2323: In summary, we can see that by turning on an external magnetic field in the 

2324: QW system, the up and down spin electrons see different renormalized strengths 

2325: of any barriers (or impurities) --- this is the odd-even effect. We speculate 

2326: on the possible use of this effect in creating a spin-valve using QW 

2327: systems. The effective actions, their resonance symmetries as well as the 

2328: temperature and length power-law corrections to the conductance in the 

2329: various temperature regimes, however, still follow a pattern similar to 

2330: that for a QW without a magnetic field. 

2331:  

2332: \section{\bf Comparison with the Experiments}

2333: 

2334: We now discuss the relevance of this model to many of the experiments that 

2335: have been performed so far on quantum wire systems fabricated using 

2336: cleaved-edge overgrowth as well as split-gate techniques. But before 

2337: doing that, let us reiterate some well-known

2338: observations about the experimental system that we are trying to 

2339: model here. In this system, the electrons enter the wire from the 

2340: 2DEG reservoirs lying outside the wire with 

2341: a Fermi energy $E_F$ whose value (typically around $5 - 10meV$) is fixed by 

2342: the parameters of the 2DEG. 

2343: Within the quantum wire, the gate voltage produces a discrete set of sub-bands 

2344: labeled by an integer $s$ (see Fig. 2); let $E_s$ denote the energies of the 

2345: bottoms of these sub-bands. In a sufficiently long quantum wire, we expect 

2346: $E_s$ to be constant along the length of the wire provided we are not too 

2347: close to either of the junctions. Thus an electron which has energy $E_F$ 

2348: and enters the sub-band $s$ will have a wave number $k_{Fs}$ inside the 

2349: wire given by $k_{Fs}^2 /2m = E_F - E_s$ and a velocity given by 

2350: $v_{Fs} = k_{Fs} /m$ \cite{buttiker}.

2351: We know that if $N$ of the $1D$ sub-bands lie below the 2DEG

2352: Fermi energy $E_{F2D}$ (which itself at any finite temperature is 

2353: surrounded by a small thermal spread), we will get 

2354: $N$ quantized steps in the conductance when the quantum wire is completely 

2355: free of any impurities; this statement is true irrespective of the electron 

2356: velocities, densities or how they interact among themselves while in the 

2357: various channels \cite{safi,maslov,buttiker}. 

2358: Now, upon increasing the gate voltage $V_G$, one adds an 

2359: energy $eV_G$ to every electron in each of the $1$D sub-bands in the quantum 

2360: wire. This has the effect of pushing up each of the sub-bands 

2361: by the same energy and can even de-populate the 

2362: sub-bands by pushing them above $E_{F2D}$ (see Fig. 2). Thus, changing 

2363: the gate voltage decreases the electron density in the quantum wire and 

2364: allows the transport process to take place through only a few channels, and 

2365: in the extreme limit, only one channel, before cutting off the wire altogether 

2366: by pushing all the $1$D sub-bands above the $E_{F2D}$ (this is called 

2367: {\it pinch-off}). The conductance measurement which 

2368: shows step quantization in terms of rises and plateaus can then be 

2369: explained in the following way. Whenever, by decreasing the 

2370: gate voltage $V_G$, the bottom one of the 1D sub-bands (which is initially 

2371: well above $E_{F2D}$) first touches the top of the thermal spread just 

2372: above $E_{F2D}$, that band starts filling up and so 

2373: we can see a rise. Once the bottom of this sub-band 

2374: crosses the bottom of the thermal spread just below $E_{F2D}$, the rise 

2375: is topped off by a plateau which signals that another channel is fully open 

2376: to electron transport between the two reservoirs (see Fig. 2). Some of the 

2377: earliest experiments with quantum wires free of impurities did indeed reveal

2378: quantization of the conductance in integer steps of $2g_0$ \cite{wees}.

2379: 

2380: But later Tarucha {\it et al} \cite{tarucha} performed experiments 

2381: with wires of lengths of $2\mu m$ to $10\mu m$ fabricated using split-gate 

2382: methods at temperatures from $0.3K$ to $1.1K$,

2383: and found deviations from the perfect quantization of the steps. Attempts

2384: were then made to explain these deviations as due to electron-electron

2385: interactions. Although, a clean TLL wire between Fermi liquid

2386: leads would not lead to renormalization of the conductance

2387: quantization, several authors \cite{furusaki,safilong,maslovlong}, showed 

2388: that the presence of impurities in a 

2389: TLL connected to Fermi leads would cause renormalization. 

2390: However, they expected the renormalizations to be gate voltage

2391: dependent; this was indeed seen by Tarucha {\it et al} \cite{tarucha}.

2392: 

2393: However, Yacoby {\it et al} \cite{yacoby} made the following surprising 

2394: observation for a quantum wire $2\mu m$ long fabricated in 

2395: cleaved-edge overgrowth systems: the dc conductance 

2396: showed several nearly flat plateaus whose 

2397: heights are uniformly renormalized from the ideal 

2398: values of integer multiples of $2g_0$ for measurements made over a 

2399: temperature range of $0.3 - 25K$. Similar observations were subsequently 

2400: made in several experiments on quantum wires made using the split-gate 

2401: technique \cite{facer,liang,liang2,reilly}. In all these experiments

2402: the step heights were increasingly renormalized as either the temperature 

2403: was lowered (for a fixed length of quantum wire) or the length of the 

2404: quantum wire was increased at a fixed temperature.

2405: Such renormalizations would require back-scattering of electrons.

2406: If these back-scatterings were due to impurities within

2407: the quantum wires, the conductance corrections would be gate

2408: voltage dependent as shown in our calculations. This can

2409: certainly not lead to flat conductance plateaus as seen in the experiments. 

2410: 

2411: Our model, however, has contact regions independent of the gate voltage and 

2412: has barriers at the contacts arising due to the changes in the

2413: nature of the electron-electron interactions and geometry. 

2414: Thus, the back-scattering at these barriers is independent of gate

2415: voltage and the sub-band index (as can be seen in our results), and will 

2416: lead to conductance plateaus which are flat as a function of 

2417: gate voltage and uniform for all the sub-bands at the highest temperatures. 

2418: We note that a recent experiment 

2419: \cite{picciotto} on a quantum wire system similar to that used by Yacoby 

2420: {\it et al} \cite{yacoby} revealed the existence of a 

2421: region of length $2 - 6\mu m$ which lies in between the gated quantum wire 

2422: region and the 2DEG reservoirs and gives rise to the back-scattering 

2423: that causes the flat and uniform renormalization of the conductance of each 

2424: sub-band. Such contact regions

2425: correspond to $T_d \sim 0.2 - 0.7 K$. This is much less than most 

2426: of the temperature range shown in Fig. 3 of Ref. \cite{yacoby}. 

2427: The similar flat and uniform conductance corrections

2428: seen in the experiments of Refs. \cite{facer,liang,liang2,reilly} 

2429: seem to suggest that their QW systems also include contact regions and have

2430: $T \gg T_d$. 

2431: 

2432: Now, as explained earlier for a quantum wire 

2433: system in which the contact length $d \ll l$, in the intermediate and 

2434: low temperature regimes of $T_l \ll T \ll T_d$ and $T \ll T_l$, 

2435: we know that the correct cutoffs for the RG procedure are 

2436: $T_d$ and $T_l$ respectively; that is why the length power-laws of 

2437: $d$ and $l$ appear in the conductance corrections in these two regimes 

2438: besides the customary temperature power-law. We can clearly see that the 

2439: inverse length scale $d^{-1}$ (for the contact region) and $l^{-1}$ (for 

2440: the wire region) have similar power-laws to those obtained for the 

2441: temperature. Thus, one 

2442: can qualitatively understand the increase in the conductance with increasing 

2443: temperature and its decrease as the length of the quantum wire is increased.

2444: This has been observed by several groups \cite{tarucha,yacoby,facer,liang2}. 

2445: Furthermore, one recent experiment using a split-gate QW system 

2446: \cite{reilly} shows that the conductance of a $2\mu m$ long QW

2447: at $T = 1K$ shows flat, renormalized plateaus which are replaced by uneven 

2448: conductance fluctuations at $T = 50mK$. However a different experiment 

2449: \cite{facer} reveals that a QPC created using similar

2450: split-gate methods shows plateaus which are hardly renormalized at higher 

2451: temperatures, and no conductance fluctuations are seen at lower temperatures. 

2452: This can also be understood from our model: the 

2453: conductance corrections due the junction barriers for a quantum wire are 

2454: gate voltage independent at higher temperatures, but are dependent on it at 

2455: lower temperatures. For the experiment in Ref. \cite{reilly}, $T_l = 0.4K >> 

2456: T=50mK$. Hence, resonance effects are expected at these temperatures. 

2457: This is in contrast to the conductance corrections for a QPC which are

2458: gate voltage independent at all temperatures. In fact, if the quantum wire 

2459: samples of Yacoby {\it et al} have contact regions as long 

2460: as $2 - 6\mu m$ (as found by the authors of Ref. \cite{picciotto} on similar 

2461: samples), this would suggest that their $2\mu m$ long wire is actually 

2462: closer to a quantum point contact. This would help explain the flatness 

2463: of the renormalizations seen over a wide temperature range of $0.3 - 25K$.

2464: 

2465: We now discuss our attempt to quantitatively understand the variation of 

2466: conductance against temperature as given in the inset of Fig. 3 of the work 

2467: of Yacoby {\it et al} \cite{yacoby}. The conductance given there is 

2468: measured at a fixed value 

2469: of the gate voltage on the plateau of the first sub-band (i.e., close to 

2470: $2g_0$). We find that the conductance correction versus temperature variation 

2471: found by them (i.e., $\delta g \equiv 2g_0 - g$ vs. $T$) is best fitted by a 

2472: function of the form

2473: \beq

2474: \delta g = - 0.3512 ~T^{-0.1058 - 0.0345 T} 

2475: \eeq

2476: as shown in Fig. 3. We find that the goodness of this fit is given by 

2477: the correlation coefficient $R^2 = 0.9955$.

2478: Clearly, this expression for the conductance corrections does not match the 

2479: simple form $\delta g \sim T^{-\alpha}$ given in Sec. $4$ for the QW or QPC 

2480: systems. The presence

2481: of the $T$ dependent piece in the exponent implies that our model is

2482: only qualitatively correct. Several factors could be important in determining 

2483: this complicated temperature dependent power-law. Some of these are:

2484: 

2485: \begin{itemize}

2486: 

2487: \item a more extended transition region between the leads and the contacts 

2488: in which the parameters $K$ and $v$ vary smoothly as a function of $x$, 

2489: 

2490: \item more extended junction barriers lying within the contacts rather than 

2491: the local $\delta$-function barriers that we have studied, and

2492: 

2493: \item the possibility of the electron-electron interactions having 

2494: a finite range instead of the short-ranged interactions that 

2495: we have used to study our TLL systems.

2496: 

2497: \end{itemize}

2498: 

2499: A detailed quantitative comparison of our model with the experiments would, 

2500: therefore, need a more sophisticated treatment taking these factors into 

2501: consideration. We should emphasize here that a temperature and length 

2502: dependence of the conductance correction of the form that we have obtained 

2503: (decreasing at high temperatures or short lengths) 

2504: is a nontrivial effect of the electron interactions, and our simple model has 

2505: already captured this qualitatively. A non-interacting theory does not have 

2506: temperature or length dependences of this kind.

2507: 

2508: We now discuss the important experimental finding of Liang {\it et al} 

2509: \cite{liang} of the odd-even effect in the transport of electrons through 

2510: a quantum wire in the presence of a magnetic field.

2511: Liang {\it et al} find that as they turn up the external magnetic field (kept 

2512: in plane and aligned along the direction of the channel) from $0$ to $11T$, 

2513: the increasing magnetic field expectedly lifts the spin degeneracy and splits 

2514: each conductance step into two steps, with the heights of both being less than 

2515: $g_0$. Furthermore, at a magnetic field strength of $11T$, they find that the 

2516: difference between the conductance of successive pairs of spin-split sub-bands 

2517: alternates. This shows that the conductance of the odd numbered 

2518: spin-split sub-bands containing the moments aligned with the magnetic field 

2519: undergoes little renormalization (i.e., is close to $g_0$ in their Fig. 4), 

2520: while the conductance of the even numbered spin-split sub-bands containing 

2521: the moments anti-aligned with the magnetic field undergoes a large 

2522: renormalization correction; their Fig. 4 indicates a correction as large 

2523: as $0.3 g_0$. As discussed earlier, this phenomenon can be simply 

2524: understood as the aligned moments seeing a much weaker barrier and the 

2525: anti-aligned moments seeing a much stronger barrier. This is due to the Zeeman 

2526: splitting of the Fermi levels of the up and down spin electrons 

2527: and their interactions with each other. 

2528: Since the difference in renormalizations between the aligned and 

2529: anti-aligned electrons occur for all magnetic fields (i.e., even when the

2530: up and down sub-bands are not spin-split), we suggest the following 

2531: possibility. One can artificially enhance the barrier

2532: strengths so that the difference in renormalizations of the up

2533: and down spins can be made substantial at moderate magnetic fields. More 

2534: importantly, we can vary the gate voltage so as to tune the 

2535: spin polarization with greater transmission to

2536: resonance. This would mean that at these values of the magnetic field

2537: and gate voltage, transmission of one of the polarizations is completely

2538: suppressed and the other one greatly enhanced. This leads us to

2539: the possibility of creating a spin-valve at moderate magnetic fields.

2540: 

2541: Finally, we comment on a new set of experiments \cite{topinka,crook} which 

2542: have used scanning probe microscopy techniques to study transport through 

2543: QPCs and propose a test for our model based on such a study. In these 

2544: experiments, a negatively charged atomic force microscope tip is held 

2545: at a distance of $100-150nm$ above the 2DEG gas on which the QPC is created 

2546: via split-gate methods. A capacitive coupling between the 2DEG and the 

2547: tip reduces the density of the 2DEG in a small spot directly beneath the tip, 

2548: thereby creating a small depletion region (negatively charged ``bubble") which 

2549: can back-scatter electrons approaching it. The tip then scans the surface 

2550: of the 2DEG reservoir into which the electrons are entering after traveling 

2551: through the QPC, and the two-probe conductance is measured. This allows one 

2552: to ``image" the electron current flowing out from the QPC. Topinka {\it et al} 

2553: \cite{topinka} have made such measurements at a temperature of $T = 1.7K$ and 

2554: find that the electrons flow out into the 2DEG reservoir in streaks from each 

2555: sub-band. The number and nature of the streaks is governed by the 

2556: electron wave function in each sub-band caused by the quantization due to the 

2557: confinement in the transverse direction. They find that the electron flow is 

2558: coherent along 

2559: these streaks quite far from the QPC mouth where they finally disperse into 

2560: the 2DEG. Furthermore, they find that placing the depletion bubble in the

2561: path of a particular streak (at a distance of about $0.3 - 0.5\mu m$ from 

2562: the mouth of the QPC) gives rise to a flat, renormalized plateau only for the 

2563: particular sub-band from which it is emanating, while the other sub-bands give 

2564: the universal conductance value of $2g_0$. This tells us that the effect 

2565: of the gate voltage must vanish quickly since it is not felt beyond distances 

2566: as short as $0.3 - 0.5\mu m$ from the mouth of the QPC. Crook {\it et al} 

2567: \cite{crook} find a series of peaks and troughs upon measuring the 

2568: differential conductance $dg/dV_G$ versus the gate voltage $V_G$ (which are 

2569: caused by the step rises and plateaus for each sub-band respectively) while 

2570: scanning the tip through the QPC. Their finding that the troughs do not fall 

2571: to zero indicate that the conductance corrections caused by the depletion 

2572: bubble (when placed within the QPC) is gate 

2573: voltage dependent as would have been expected. 

2574: 

2575: Now, the availability of the tip generated depletion bubble as a controlled 

2576: barrier to the flow of electrons through the QPC also 

2577: suggests a possible use of scanning probe microscopy 

2578: techniques to test the predictions of 

2579: our model in a quantitative fashion. This would require the gate voltage to be 

2580: first fixed such that only the lowest sub-band is fully open to the flow 

2581: of electron current, and then the depletion bubble to be placed somewhere on a 

2582: streak emanating from this lowest sub-band at a distance from the QPC mouth; 

2583: the conductance can then be measured by changing the gate 

2584: voltage but holding the temperature fixed. The nature of the 

2585: conductance versus gate voltage curve will tell us whether the gate voltage 

2586: does or does not have any effect on the electrons on the streak at that 

2587: distance from the mouth of the QPC. Furthermore, the gate voltage can then be 

2588: held fixed somewhere on a plateau and the conductance measured as the 

2589: temperature is varied. The form of the conductance corrections versus 

2590: temperature can then be obtained. This entire chain of measurements can then 

2591: be repeated after taking the depletion bubble closer to the QPC mouth and 

2592: into the QPC in a series of steps. Such a series of measurements would 

2593: help answer questions about where the conductance corrections start becoming 

2594: dependent on the gate voltage as well as 

2595: how the conductance corrections vary with temperature when a barrier is 

2596: placed within the QPC or away from the QPC. Such 

2597: experiments could also be carried out with longer QWs to 

2598: check the length dependences of the conductance corrections.

2599: 

2600: \section{\bf Summary and Outlook}

2601: 

2602: The main idea in this paper is to introduce a model which explicitly

2603: describes the regions in between the quantum wire and the 2DEG

2604: reservoirs as interacting 1D systems which

2605: are independent of the density of electrons in the quantum wire.

2606: We show that the difference in the strengths of the interactions in

2607: the different regions leads to local junction barriers between the regions; 

2608: the barriers simulate the effects of the imperfect coupling between 

2609: the 2DEG and the quantum wire. Our model leads to the following results for 

2610: wires with no impurities, all of which are in agreement with a large body of 

2611: experimental observations.

2612: 

2613: \begin{itemize}

2614: 

2615: \item{} Flat (independent of gate voltage) and uniform (for all the sub-bands) 

2616: renormalizations of the quantized conductance plateaus.

2617: 

2618: \item{} The renormalizations increase as the temperature is 

2619: lowered or the length of the quantum wire is increased.

2620: 

2621: \item{} At still lower temperatures, the flatness of the plateaus

2622: disappears and oscillatory features in the conductance can be observed 

2623: which we interpret as resonant transmission through the quantum wire.

2624: 

2625: \item{} In the presence of a magnetic field, an odd-even effect

2626: is found in the conductance of alternate spin-split sub-bands. This effect 

2627: may be used to construct a spin-valve, which allows only electrons with one 

2628: particular spin to transmit through the wire even if the magnetic field is 

2629: not high enough to completely spin-split the sub-bands.

2630:  

2631: \end{itemize} 

2632: 

2633: For quantum wires with impurities, which are either intrinsic or

2634: externally imposed as finger gates, the conductance corrections

2635: are always gate voltage dependent and therefore, are neither flat

2636: nor sub-band independent. 

2637: 

2638: Some interesting questions for future studies include the following.

2639: A quantitative fit to the conductance corrections as a function

2640: of the temperature and wire length still remains to be done. This

2641: would require an even more realistic modeling of the quantum wire

2642: system (including some of the features itemized in the

2643: previous section) as well as more experimental data. 

2644: Theoretical studies at finite frequencies and finite external voltages

2645: across the quantum wire also need to be pursued. 

2646: Finally, one needs to understand several features which are observed

2647: on the rise between two successive plateaus, such as the ``$0.7$ effect''

2648: mentioned in the introduction, the observation of continuous oscillations as 

2649: a function of the gate voltage upon introducing finger gate barriers

2650: \cite{tkachenko}, and the fixed point that exists on the rise as the 

2651: temperature is varied \cite{senz}. For all of these, one needs to study the 

2652: model when some sub-band is partially opened.

2653:  

2654: \vskip .5 true cm

2655: \leftline{\bf Acknowledgments}

2656: \vskip .5 true cm

2657: 

2658: SL thanks I. Safi for useful correspondence. DS thanks the Council of 

2659: Scientific and Industrial Research, India for financial support through 

2660: grant No. 03(0911)/00/EMR-II.

2661: 

2662: \vskip .5 true cm

2663: 

2664: \appendix

2665: 

2666: \section{\bf Effective action for spinless fermions}

2667: 

2668: In this Appendix, we will obtain explicitly the $S_0$ part of the effective 

2669: action in terms of the fields $\phi_i,i=1,..4$, at the junctions $x=0,d,l+d$ 

2670: and $x=l+2d=L$ for the $K_L$-$K_C$-$K_W$-$K_C$-$K_L$ model described by the 

2671: action in Eq. (\ref{s0}) by integrating out all degrees of freedom except at 

2672: the positions of the junctions. We will also give the effective action of the

2673: simpler model $K_L$-$K_W$-$K_L$ for comparison, since they have also

2674: not been explicitly given anywhere.

2675: 

2676: We first start with the simpler $K_L$-$K_W$-$K_C$ model, which is 

2677: defined as a length $L$ quantum wire with interaction parameter $K_W$ 

2678: between $x=0$ and $x=L$, and with leads defined by $K_L=1$ for

2679: $x<0$ and $x>L$, described by the Lagrangian 

2680: \beq

2681: {\cal L}= \int_{-\infty}^0 dx {\cal L}_1 + \int_{0}^{L} 

2682: dx {\cal L}_2 + \int_L^\infty dx {\cal L}_1 ,

2683: \eeq

2684: where

2685: \bea

2686: {\cal L}_1 ~&=&~ {1\over 2K_Lv_L} (\pt\phi)^2 +{v_L\over 2K_L} 

2687: (\px\phi)^2, \nonum \\

2688: {\rm and } \quad 

2689: {\cal L}_2 ~&=&~ {1\over 2K_Wv_W} (\pt\phi)^2 +{v_W\over 2K_W} (\px \phi)^2.

2690: \eea

2691: to set the notation. There 

2692: are three ways to derive the effective action. We can (a) integrate out

2693: the fields at all points in space except at $x=0$ and $L$, or (b) find 

2694: the solution of the equations of motion in terms of the above two fields

2695: and then compute the action from that solution, or (c) compute the 

2696: Green's function $G_\bo (x,x^\prime )$, set $x,x^\prime$ equal

2697: to $0$ or $L$, and invert $G$ to get $S_{eff}$. All the methods 

2698: produce the same result since the original action is purely quadratic.

2699: We will use the second method here because it is technically simpler.

2700: 

2701: As in other sections, we will work with the 

2702: Euclidean time action for convenience. If all the

2703: fields have a time dependence of the form $\exp (-i\on \tau)$, then 

2704: normalizability of the solutions imply that they should decay exponentially

2705: at $x \rightarrow \pm \infty$. We assume that the solution of the equation

2706: of motion has the following forms in the three regions,

2707: \bea

2708: \tphi(x,\on) &=& \tphi(0,\on)e^{-i\on\tau +|\on|x/v_L}, \quad x<0 \nonum \\

2709: &=& e^{-i\on\tau} (\ttheta_1(\on) e^{|\on| x/v_W} +\ttheta_2(\on)

2710: e^{-\on x/v_W}), \quad 0<x<L \nonum \\

2711: &=& \tphi(L,\on)e^{-i\on\tau+|\on|(L-x)/v_L}, \quad x>L ~.

2712: \eea

2713: Matching solutions at the boundaries $x=0$ and $x=L$ to eliminate $\ttheta_i

2714: (\on)$, and using this solution in the effective action and carrying out the 

2715: spatial integration, we obtain the action 

2716: \bea

2717: S_0 &=& {1\over 2K_L} \sum_{\on} |\on| (\tphi_1^2+\tphi_2^2) 

2718: + {1\over 2K_W} \sum_{\on} {|\on| \over e^{k_{nW}L} - e^{-k_{nW}L} } 

2719: \times \nonum \\ 

2720: && \quad \quad \quad [(e^{k_{nW}L}+e^{-k_{nW}L} )(\tphi_1^2+\tphi_2^2) -

2721: 4\tphi_1\tphi_2] ,

2722: \eea

2723: where $\tphi_1 \equiv \tphi(0,\on)$ and $\tphi_2 \equiv \tphi(L,\on)$

2724: and $k_{nW}$ and $k_{nL}$ are defined as $|\on|/v_W$ and $|\on|/v_L$ 

2725: respectively. 

2726: In the limit $\on \gg v_L/L, v_W/L$, we get the high frequency effective action 

2727: \beq

2728: S_{high} = {K_L+K_W \over 2K_LK_W} \sum_{\on}|\on|

2729: (\tphi_1^2 +\tphi_2^2) ~,

2730: \eeq

2731: where the two junctions are decoupled as expected. In the low

2732: frequency limit $\on \ll v_L/L, v_W/L$, we get

2733: \beq

2734: S_{low} = {1\over 2K_L} \sum_{\on} |\on| (\tphi_1^2+

2735: \tphi_2^2) + { U_W\over 2} \sum_{\on} |\on| (\tphi_1-\tphi_2)^2 ~,

2736: \eeq

2737: where $U_W = v_W/K_W L$. 

2738: 

2739: Using the same method as above, we can also obtain the full effective

2740: action for the $K_L$-$K_C$-$K_W$-$K_C$-$K_L$ model in terms of the fields at

2741: the four junctions $\phi_i, i=1,..4$, where $\tphi(0,\on) = \tphi_1(\on) 

2742: \equiv \tphi_1$, $\tphi(d,\on) \equiv \tphi_2$, $\tphi(l+d,\on)

2743: \equiv \tphi_3$ and $\tphi(L=l+2d,\on) \equiv

2744: \tphi_4$. The solutions in the five regions can be written as

2745: \bea

2746: \tphi(x, \on) &=& \tphi_1 e^{k_{nL}x}, \quad x<0 \nonum \\

2747: &=& B e^{k_{nC}x} + Ce^{-k_{nC}x}, \quad 0<x<d \nonum \\

2748: &=& D e^{k_{nW}x} + Ee^{-k_{nW}x}, \quad d<x<l+d \nonum \\

2749: &=& F e^{k_{nC}x} + Ge^{-k_{nC}x}, \quad l+d<x<L \nonum \\

2750: &=& \tphi_4 e^{k_{nL}(L-x)}, \quad x<0 ~.

2751: \eea

2752: where by matching the solutions at $x=0,d,l+d$ and $L$, we can

2753: obtain the functions of $\on$, $B,C,D,E,F$ and $G$ in terms of the

2754: $\tphi_i, i=1...4$, and $k_{nC}$ is defined as $k_{nC} = |\on| /v_C$. 

2755: Substituting this solution in the 

2756: action and integrating over all space, we get the effective action

2757: \bea

2758: S_{eff} &=& \sum_{\on}\{{1\over 2K_L} |\on| (\tphi_1^2+\tphi_4^2) \nonum \\

2759: && +{1\over 2K_C} |\on| [(B^2+G^2e^{-2k_n L}) (e^{2k_n d} - 1) + (C^2+F^2

2760: e^{2k_nL})(1-e^{-2k_n d})] \nonum \\

2761: && +{1\over 2K_W} [|\on| [D^2(e^{2k_n (L-d)} - e^{2k_n d}) -

2762: E^2(e^{-2k_n (L-d)} - e^{-2k_n d})] \}~,

2763: \eea

2764: with

2765: \bea

2766: B={\tphi_2 -\tphi_1e^{-k_{nC}d}\over e^{k_{nC} d} -e^{-k_{nC}d}}, \quad && C=

2767: {\tphi_1e^{k_{nC}d}-\tphi_2\over e^{k_{nC} d} -e^{-k_{nC}d}}\nonum \\

2768: D={\tphi_3 -\tphi_2e^{-k_{nW}l}\over e^{k_{nC}d} (e^{k_{nW}l} -e^{-k_{nW}l})}, 

2769: \quad && E= {\tphi_2e^{k_{nW}l}-\tphi_2\over e^{-k_{nC}d}

2770: (e^{k_{nW} l} -e^{-k_{nW}l})}\nonum \\

2771: F={\tphi_4 -\tphi_3e^{-k_{nC}d}\over e^{k_{nC}d +k_{nW}l}

2772: (e^{k_{nC} d} -e^{-k_{nC}d})}, 

2773: \quad && G= {\tphi_3e^{k_{nC}d}-\tphi_4\over e^{-k_{nC}d -k_{nW}l}

2774: (e^{k_{nC} d} -e^{-k_{nC}d})} ~. 

2775: \eea

2776: The action finally simplifies to 

2777: \bea

2778: S_{eff} &=& {1\over 2K_L} \sum_{\on} |\on| (\tphi_1^2+\tphi_4^2) \nonum \\

2779: &&\hspace*{-0.2cm} +{1\over 2K_C} \sum_{\on} {|\on| \over e^{k_{nC}d} 

2780: - e^{-k_{nC}d}} [(e^{k_{nC}d} +e^{-k_{nC}d}) (\tphi_1^2 +

2781: \tphi_2^2 + \tphi_3^2+\tphi_4^2) -4\tphi_1\tphi_2 -4\tphi_3\tphi_4] \nonum \\

2782: &&\hspace*{-0.2cm} +{1\over 2K_W} \sum_{\on} {|\on| \over e^{k_{nW}l} 

2783: - e^{-k_{nW}l}} [(e^{k_{nW}l} +e^{-k_{nW}l}) (\tphi_2^2+\tphi_3^2) 

2784: -4\tphi_2\tphi_3] ~,

2785: \eea

2786: where $\tphi_1 \equiv \tphi(0,\on)$ and $\tphi_2 \equiv \tphi(L,\on)$.

2787: The high and low frequency limits of this effective

2788: action have been used in Sec. 3, to compute the finite temperature and

2789: finite length corrections off-resonance.

2790: 

2791: \section{\bf Effective action for spinful fermions}

2792: 

2793: In this section, we will explicitly compute the effective action for 

2794: spinful fermions in the $K_L$-$K_C$-$K_W$-$K_C$-$K_L$ model. Although,

2795: the method followed is exactly the same as that in the previous

2796: section for spinless fermions, we do it explicitly because there are 

2797: a few points where the inclusion of spin makes a difference.

2798: 

2799: The effective action for spinful fermions is normally computed in

2800: terms of the `charge' and `spin' field variables defined as $\phi_\rho= 

2801: (\phi_\ua +\phi_\da)/\sqrt{2}$ and $\phi_\sigma =(\phi_\ua

2802: -\phi_\da)/\sqrt{2}$ because in the presence of interactions, the spin $\ua$

2803: and $\da$ fermions are mixed (remember the Hubbard term ${\cal U} \sum_i

2804: n_{i\ua} n_{i\da}$). Here, in our model with contacts, the interaction

2805: term $\cal U$ is different in the contact region and in the wire

2806: region. But since the linear combination that diagonalizes the interaction 

2807: is independent of the value of $\cal U$, the action in terms of the 

2808: $\phi_\rho$ and $\phi_\sigma$ fields are decoupled. In the 

2809: presence of a magnetic

2810: field, in the next Appendix, we will see that the action continues to be

2811: diagonalizable; however, the diagonal fields are defined in terms of

2812: mixing angles which explicitly depend on $\cal U$ and the magnetic

2813: field and hence are different in the leads, the contacts and the wire. 

2814: 

2815: The starting action for the spinful fermions is given in

2816: Eqs. (\ref{s0}) and (\ref{sol}) in the text in terms of the charge and

2817: spin fields. As in the earlier Appendix, we will obtain the solution of the 

2818: equations of motion in terms of the eight fields $\tphi_{ia}, i=1...4,

2819: a=\rho , \sigma$ defined to be at the positions $x=0,d,l+d$ and $L=l+2d$ and

2820: then compute the effective action from that solution. We 

2821: assume that the solutions in the five regions can be written as

2822: \bea

2823: \tphi_{a}(x, \on) &=& \tphi_{1a} e^{k_{nLa}x}, \quad x<0 \nonum \\

2824: &=& B_a e^{k_{nCa}x} + C_ae^{-k_{nCa}x}, \quad 0<x<d \nonum \\

2825: &=& D_a e^{k_{nWa}x} + E_ae^{-k_{nWa}x}, \quad d<x<l+d \nonum \\

2826: &=& F_a e^{k_{nCa}x} + G_ae^{-k_{nCa}x}, \quad l+d<x<L \nonum \\

2827: &=& \tphi_{4a} e^{k_{nL}(L-x)}, \quad x<0 ~.

2828: \eea

2829: and as before, the coefficients, $B_a,C_a,D_a,E_a,F_a$ and $G_a$ can

2830: be found in terms of the $\tphi_{ia}, i=1...4, a=\rho , \sigma$

2831: by matching the solutions at $x=0,d,l+d$ and $L$. Note that 

2832: $k_{nWa}$, $k_{nCa}$ and $k_{nL}$ are defined as $|\on|/v_{Wa}$, 

2833: $|\on|/v_{Ca}$ and $|\on|/v_{La}$ respectively. Substituting this solution 

2834: in the action and integrating over all space, we get the effective action as

2835: \bea

2836: S &=& \sum_{a=\rho , \sigma}\{{1\over 2K_{La}} 

2837: \sum_{\on} |\on| (\tphi_{1a}^2+\tphi_{4a}^2) + {1\over 2K_{Ca}} 

2838: \sum_{\on} {|\on| \over e^{k_{nCa}d} - e^{-k_{nCa}d}} \times\nonum \\ 

2839: && \quad \quad \quad [(e^{k_{nCa}d} +e^{-k_{nCa}d}) (\tphi_{1a}^2+

2840: \tphi_{2a}^2 + \tphi_{3a}^2+\tphi_{4a}^2) -4\tphi_{1a}\tphi_{2a}

2841: -4\tphi_{3a}\tphi_{4a}] \nonum \\

2842: &+& {1\over 2K_{Wa}} \sum_{\on} {|\on| \over e^{k_{nWa}l} - e^{-k_{nWa}l}} 

2843: [(e^{k_{nWa}l} +e^{-k_{nWa}l}) (\tphi_{2a}^2+\tphi_{3a}^2)

2844: -4\tphi_{2a} \tphi_{3a}]\}.

2845: \label{spineff} 

2846: \eea

2847: The high and low frequency limits of this effective

2848: action have been used in Sec. 3, to discuss the various resonances that

2849: are possible in the low temperature limit and to explicitly 

2850: compute the off-resonance corrections to the conductances at 

2851: finite temperatures and for finite length wires.

2852: 

2853: \section{\bf Effective action for spinful electrons in the

2854: presence of a magnetic field}

2855: 

2856: We present here the calculation for the effective action for

2857: spinful fermions in a magnetic field 

2858: when the mixing angle $\gamma$ is the same in both the contacts as well as 

2859: the QW, i.e., the short ranged electron-electron interaction $\cal U$ is equal 

2860: in all the three TLLs. We start with the action given in Eq. (\ref{actmag}) 

2861: in Sec. 5 and will now integrate out the fields at all points except at the

2862: four junctions as these will be 

2863: the sites for the two outer barriers while the two inner junctions are the 

2864: ends of the region to which the gate voltage couples. Thus, we write down 

2865: the equations of motion in each of the five regions and solve them. If all the

2866: fields have a time dependence of the form $\exp (-i\on \tau)$, then 

2867: normalizability of the solutions imply that they should decay exponentially

2868: at $x \rightarrow \pm \infty$. The general solution is given by

2869: \bea

2870: \tphi_\ua(x) &=& A_\ua e^{k_\ua x}, \quad \tphi_\da(x) = A_\da e^{k_\da x}, 

2871: \quad x<0 \nonum \\

2872: \tphi_\pm(x) &=& B_{\pm}e^{k_\pm x}+C_{\pm}e^{-k_\pm x}, \quad 0<x<d \nonum \\

2873: \tphi_\pm(x) &=& D_{\pm}e^{\tk_\pm x}+E_{\pm}e^{-\tk_\pm x}, \quad d<x<l+d 

2874: \nonum \\ 

2875: \tphi_\pm(x) &=& F_{\pm}e^{k_\pm x}+G_{\pm}e^{-k_\pm x}, \quad l+d<x<l+2d 

2876: \nonum \\

2877: \tphi_\ua(x) &=& H_\ua e^{k_\ua x}, \quad \tphi_\da = H_\da e^{k_\da x}, \quad

2878: x<0 ~,

2879: \label{coeff}

2880: \eea

2881: where we have defined $\tk_\pm = |\on| /v_{W\pm}$, $k_\pm = |\on| /

2882: v_{C\pm}$, $k_\ua = |\on|/v_{F\ua}$ and $k_\da = |\on|/v_{F\da}$.

2883: 

2884: We now solve for the coefficients $A,...,H$ in Eq. (\ref{coeff})

2885: by matching the fields

2886: $\tphi_\ua$ and the $\tphi_\da$ at the four junctions. At this point, we make

2887: the simplifying assumption that the mixing angles $\gamma_C$ and $\gamma_W$

2888: (defined as in Eqs. (\ref{transf1}) and (\ref{transf2})) in the contact and 

2889: wire regions are equal to each other, $\gamma_C = \gamma_W = \gamma$. 

2890: This implies that $\tphi_{W\pm} = \sqrt{{K_{W\pm} v_{W\pm}} \over {K_{C\pm} 

2891: v_{C\pm}}} \tphi_{C\pm}$ at $x=d$ and $x=l+d$.

2892: We find that the coefficients are given by 

2893: \bea

2894: A_\ua &=& \tphi_{1\ua}, \quad A_\da = \tphi_{1\da} \nonum \\

2895: B_{\pm} &=& {\tphi_{2\pm} - \tphi_{1\pm}

2896: e^{-k_\pm d}\over D_{C\pm}} \nonum \\

2897: C_{\pm} &=&- {\tphi_{2\pm} + \tphi_{1\pm}

2898: e^{k_\pm d}\over D_{C\pm}} \nonum \\

2899: D_{\pm} &=& \sqrt{{K_{W\pm} v_{W\pm}} \over {K_{C\pm} v_{C\pm}}}

2900: e^{-\tk_\pm d} ~{(-\tphi_{2\pm} 

2901: e^{-\tk_\pm l} + \tphi_{3\pm}) \over D_{W\pm}} 

2902: \nonum \\

2903: E_{\pm} &=& \sqrt{{K_{W\pm} v_{W\pm}} \over {K_{C\pm} v_{C\pm}}} 

2904: e^{\tk_\pm d}~{(\tphi_{2\pm} 

2905: e^{\tk_\pm l} - \tphi_{3\pm}) \over D_{W\pm}} 

2906: \nonum \\

2907: F_{\pm} &=& -{\tphi_{3\pm}e^{-k_\pm(l+2d)} - \tphi_{4\pm} 

2908: e^{-k_\pm d}\over D_{C\pm}} \nonum \\

2909: G_{\pm} &=& {\tphi_{3\pm}e^{k_\pm(l+2d)} - \tphi_{4\pm}

2910: e^{k_\pm d}\over D_{C\pm}} \nonum \\

2911: H_{\ua} &=& \tphi_{4\ua}e^{k_\ua(l+2d)}, \quad H_\da = \tphi_{4\da}

2912: e^{k_\da(l+2d)} ~,

2913: \eea

2914: where

2915: \bea

2916: D_{C\pm} &=& e^{k_\pm d} - e^{-k_\pm d} ~ \nonum \\

2917: {\rm and}\quad D_{W\pm} &=& e^{\tk_\pm l} - e^{-\tk_\pm l} ~.

2918: \eea

2919: 

2920: Then using the relations written down in Eq. (\ref{transf1})

2921: connecting the $\phi_{\pm}$ and $\phi_{\ua,\da}$ fields, 

2922: we find that the effective Lagrangian density is given by

2923: \bea

2924: {\cal L} = && \frac{|\on|}{2K_L}(\tphi_{1\ua}^2 + \tphi_{1\da}^2 

2925: \tphi_{4\ua}^2 + \tphi_{4\da}^2) \nonum \\

2926: && + \frac{|\on|}{2}[(p^2\frac{N_{C+}} {K_{C+}D_{C+}} + r^2\frac{N_{C-}}

2927: {K_{C-}D_{C-}})(\tphi_{1\ua}^2 + \tphi_{2\ua}^2 

2928: + \tphi_{3\ua}^2 + \tphi_{4\ua}^2) \nonum \\

2929: && + (q^2\frac{N_{C+}} {K_{C+}D_{C+}} + s^2\frac{N_{C-}}

2930: {K_{C-}D_{C-}})(\tphi_{1\da}^2 

2931: + \tphi_{2\da}^2 + \tphi_{3\da}^2 + \tphi_{4\da}^2) \nonum \\

2932: && - 4(\frac{p^2}{K_{C+}D_{C+}} + 

2933: \frac{r^2}{K_{C-}D_{C-}})(\tphi_{1\ua}\tphi_{2\ua} 

2934: + \tphi_{3\ua}\tphi_{4\ua}) \nonum \\

2935: && - 4(\frac{q^2}{K_{C+}D_{C+}} + \frac{s^2}{K_{C-}D_{C-}})

2936: (\tphi_{1\da} \tphi_{2\da} + \tphi_{3\da}\tphi_{4\da})] \nonum \\

2937: && + |\on| [(\frac{pq N_{C+}}{K_{C+} D_{C+}} 

2938: + \frac{rs N_{C-}}{K_{C-} D_{C-}})

2939: (\tphi_{1\ua}\tphi_{1\da} + \tphi_{2\ua}\tphi_{2\da} 

2940: + \tphi_{3\ua}\tphi_{3\da} + \tphi_{4\ua}\tphi_{4\da}) \nonum \\

2941: && - 2(\frac{pq}{K_{C+}D_{C+}} + \frac{rs}

2942: {K_{C-}D_{C-}})(\tphi_{1\ua}\tphi_{2\da} + 

2943: \tphi_{1\da}\tphi_{2\ua} + \tphi_{3\ua}\tphi_{4\da} 

2944: + \tphi_{3\da}\tphi_{4\ua})] \nonum \\

2945: && + \frac{|\on|}{2}[(p^2\frac{N_{W+}} {K_{W+}D_{W+}} + r^2\frac{N_{W-}}

2946: {K_{W-}D_{W-}})(\tphi_{2\ua}^2 + \tphi_{3\ua}^2) \nonum \\

2947: && + (q^2\frac{N_{W+}} {K_{W+}D_{W+}} + s^2\frac{N_{W-}}

2948: {K_{W-}D_{W-}})(\tphi_{2\da}^2 + \tphi_{3\da}^2) \nonum \\

2949: && - 4(\frac{p^2}{K_{W+}D_{W+}} +

2950: \frac{r^2}{K_{W-}D_{W-}})\tphi_{2\ua}\tphi_{3\ua}

2951: \nonum \\

2952: && - 4(\frac{q^2}{K_{W+}D_{W+}} + 

2953: \frac{s^2}{K_{W-}D_{W-}})\tphi_{2\da}\tphi_{3\da}] \nonum \\

2954: && + |\on| [(\frac{pq N_{W+}}{K_{W+} D_{W+}} + \frac{rs N_{W-}}{K_{W-}D_{W-}})

2955: (\tphi_{2\ua}\tphi_{2\da} + \tphi_{3\ua}\tphi_{3\da}) \nonum \\

2956: && - 2(\frac{pq}{K_{W+}D_{W+}} + \frac{rs}

2957: {K_{W-}D_{W-}})(\tphi_{2\ua}\tphi_{3\da} + \tphi_{2\da}\tphi_{3\ua})],

2958: \eea

2959: where

2960: \bea

2961: N_{C\pm} &=& e^{k_\pm d} + e^{-k_\pm d} ~, \nonum \\

2962: N_{W\pm} &=& e^{\tk_\pm l} + e^{-\tk_\pm l} ~,

2963: \eea

2964: and $D_{C\pm}, D_{W\pm}, k_{\pm}$ and $\tk_{\pm}$ have already been 

2965: defined above.

2966: 

2967: \section{\bf Calculation of the Green's function in our model for the 

2968: Quantum Wire}

2969: 

2970: Here, we will present a calculation of the Green's function for the bosonic 

2971: excitations in the model that we have presented for the quantum wire system 

2972: of spinless fermions. The method we follow is along the lines of the 

2973: calculation presented by Maslov and Stone \cite{maslov}. We will study 

2974: the case when there are no barriers present anywhere in the system. Then, 

2975: we see that the Euclidean action $S_E$ in all the five distinct TLL regions 

2976: in our model (Fermi lead, contact, QW, contact and Fermi lead) is given by

2977: \beq

2978: S_E = \frac{1}{2} \int_{0}^{\beta} d\tau \int_{-\infty}^{\infty}

2979: [\frac{1}{K(x)v(x)}(\partial_{\tau}\phi)^2 + 

2980: \frac{v(x)}{K(x)}(\partial_{x}\phi)^2],

2981: \eeq

2982: with $K(x) = K_L, v(x) = v_L$ in the first and fifth (Fermi lead) regions, 

2983: $K(x) = K_C, v(x) = v_C$ in the second and fourth (contact) regions and 

2984: $K(x) = K_W, v(x) = v_W$ in the third (QW) region. Then, defining the 

2985: two-point bosonic Green's function/propagator (in Euclidean time $\tau$) as

2986: \beq

2987: G(x,x',\tau) = <T_{\tau}\phi(x,\tau)\phi(x',0)>,

2988: \eeq

2989: it can be shown that the equation satisfied by the Fourier transform of 

2990: the above Green's function $G_{\bar{\omega}}(x,x')$ is

2991: \beq

2992: \{-\partial_{x}(\frac{v(x)}{K(x)}\partial_{x}) + 

2993: \frac{1}{K(x)v(x)}\bar{\omega}^2\} G_{\bar{\omega}}(x,x') = \delta(x-x').

2994: \eeq

2995: We now have to solve the above equation to obtain a functional form for 

2996: $G_{\bar{\omega}}(x,x')$. We know that the interaction parameter $K$ and 

2997: the velocity $v$ change abruptly at each of the junctions and that the 

2998: two Fermi leads are semi-infinite in length (i.e. $G_{\bar{\omega}}(x,x')$ 

2999: must decay to zero as $x\rightarrow \pm\infty$). As we are interested in 

3000: finding the one-point Green's function at a point in the left contact, we 

3001: will choose $x'$ to lie between $0$ (the left lead-contact junction) and 

3002: $d$ (the left contact-QW junction). Furthermore, we know that the 

3003: Green's function $G_{\bar{\omega}}(x,x')$ must satisfy the following 

3004: boundary conditions: (a) $G_{\bar{\omega}}(x,x')$ must be continuous at 

3005: $x = 0, x', d, l+d$ and $l+2d$ (b) $(\frac{v(x)}{K(x)})\partial_{x}

3006: G_{\bar{\omega}}(x,x')$ must be continuous at $x = 0, d, l+d$ and $l+2d$ 

3007: and 

3008: \beq

3009: - \frac{v(x)}{K(x)}\partial_{x}G_{\bar{\omega}}(x,x')

3010: \vert_{x=x'-0}^{x=x'+0} = 1,

3011: \eeq

3012: i.e., $(\frac{v(x)}{K(x)})\partial_{x}G_{\bar{\omega}}(x,x')$ undergoes a 

3013: jump of unity at $x=x'$. It is then easily seen that the solution for 

3014: $G_{\bar{\omega}}(x,x')$ is of the form

3015: \bea

3016: G_{\bar{\omega}}(x,x') &=& A e^{\vert\bar{\omega}\vert x/v_L} 

3017: \quad \quad \quad \quad ~~~~~~~~~\textrm{for $x\leq 0$} \nonum \\

3018: &=& B e^{\vert\bar{\omega}\vert x/v_C} + 

3019: C e^{-\vert\bar{\omega}\vert x/v_C} \quad ~\textrm{for $0<x\leq x'$} \nonum \\

3020: &=& D e^{\vert\bar{\omega}\vert x/v_C} + 

3021: E e^{-\vert\bar{\omega}\vert x/v_C} \quad ~\textrm{for $x'<x\leq d$} \nonum \\

3022: &=& F e^{\vert\bar{\omega}\vert x/v_C} + 

3023: G e^{-\vert\bar{\omega}\vert x/v_C} \quad ~\textrm{for $d<x\leq l+d$} \nonum \\

3024: &=& H e^{\vert\bar{\omega}\vert x/v_C} + 

3025: I e^{-\vert\bar{\omega}\vert x/v_C} \quad ~~\textrm{for $l+d<x\leq l+2d$} 

3026: \nonum \\

3027: &=& J e^{-\vert\bar{\omega}\vert x/v_C} \quad \quad \quad \quad 

3028: ~~~~~~~~\textrm{for $l+2d<x$} ~.

3029: \eea 

3030: The coefficients $A, B, \ldots , J$ are found by matching the boundary 

3031: conditions. To begin with, it is worth noting that in the dc limit 

3032: of $\bar{\omega}\rightarrow 0$, we find that 

3033: \beq

3034: A = B + C = D + E = F + G = H + I = J = \frac{K_L}{2\vert\bar{\omega}\vert},

3035: \eeq

3036: which gives the dc conductance to be

3037: \beq

3038: g = \frac{2e^2}{h} K_L = 2g_0 K_L.

3039: \eeq

3040: This gives the perfect quantized conductance observed in several experiments 

3041: on transport of electrons through a QPC when we take 

3042: the leads to be Fermi liquids with $K_L = 1$. 

3043: 

3044: We now give the expressions for the Green's functions for the case when 

3045: both $x$ and $x'$ are taken equal to $a$ at a point in the left contact:

3046: \bea

3047: &&\hspace{-1cm}G_{\bar{\omega}}(a,a)= \nonum \\

3048: &&\hspace{-1cm}\frac{K_C}{2\vert\bar{\omega}\vert}

3049: \frac{\{(r-p-q-s)e^{\vert\bar{\omega}\vert d/v_C}(1 + 

3050: \gamma_1 e^{-\vert\bar{\omega}\vert 2a/v_C}) - 

3051: (r+p+q-s)e^{-\vert\bar{\omega}\vert d/v_C}

3052: (e^{\vert\bar{\omega}\vert 2a/v_C} + \gamma_1)\}}

3053: {\{(r-p-q-s)e^{\vert\bar{\omega}\vert d/v_C} + 

3054: \gamma_1 (r+p+q-s)e^{-\vert\bar{\omega}\vert d/v_C}\}} , \nonum \\

3055: &&

3056: \eea

3057: where

3058: \bea

3059: p &=& e^{-\vert\bar{\omega}\vert l(\frac{1}{v_C}+\frac{1}{v_W})}

3060: (\gamma_1 e^{-2\vert\bar{\omega}\vert d/v_C} (1+\frac{K_W}{K_C}) 

3061: + (1-\frac{K_W}{K_C})) \nonum \\

3062: q &=& e^{-\vert\bar{\omega}\vert l(\frac{1}{v_C}-\frac{1}{v_W})}

3063: (\gamma_1 e^{-2\vert\bar{\omega}\vert d/v_C} (1-\frac{K_W}{K_C}) 

3064: + (1+\frac{K_W}{K_C})) \nonum \\

3065: r &=& e^{-\vert\bar{\omega}\vert l(\frac{1}{v_C}+\frac{1}{v_W})}

3066: (\gamma_1 e^{-2\vert\bar{\omega}\vert d/v_C} (1+\frac{K_C}{K_W}) 

3067: + (\frac{K_C}{K_W}-1)) \nonum \\

3068: s &=& e^{-\vert\bar{\omega}\vert l(\frac{1}{v_C}-\frac{1}{v_W})}

3069: (\gamma_1 e^{-2\vert\bar{\omega}\vert d/v_C} (\frac{K_C}{K_W}-1) 

3070: + (1+\frac{K_C}{K_W})) \nonum \\

3071: \gamma_1 &=& \frac{K_L - K_C}{K_L + K_C}.

3072: \eea

3073: The results upon taking the limits corresponding to the various 

3074: frequency (or temperature) regimes are given in the section 

3075: where the conductance is computed for quantum wires and quantum 

3076: point contacts with a junction barrier in the left contact region and we

3077: will not repeat them here. 

3078: 

3079: We also give the general form of the two-point propagator 

3080: $G_{\bar{\omega}}(x,y)$ for when $x$ is a point in the right lead 

3081: and $y$ is a point in the left contact: 

3082: \beq

3083: G_{\bar{\omega}}(x,y) = \frac{2K_2}{\vert\bar{\omega}\vert} 

3084: \frac{e^{\vert\bar{\omega}\vert (l+2d)(\frac{1}{v_L}-\frac{1}{v_C})} 

3085: e^{\vert\bar{\omega}\vert \frac{d}{v_C}} 

3086: (1+\gamma_1)^{2} e^{\vert\bar{\omega}\vert (\frac{y}{v_C}- \frac{x}{v_L})}}

3087: {\{(p+q+s-r)e^{\vert\bar{\omega}\vert d/v_C} +

3088: \gamma_1 (s-r-p-q)e^{-\vert\bar{\omega}\vert d/v_C}\}} ~,

3089: \eeq

3090: where the expressions for $p, q, r, s$ and $\gamma_1$ have already been 

3091: given earlier.

3092: 

3093: Now, we give the expression for the one-point propagator at a 

3094: point $a$ inside the quantum wire:

3095: \beq

3096: G_{\bar{\omega}}(a,a)= \frac{K_W}{2\vert\bar{\omega}\vert}

3097: \frac{(je^{\vert\bar{\omega}\vert a/v_W} + 

3098: ke^{-\vert\bar{\omega}\vert a/v_W}) (me^{\vert\bar{\omega}\vert a/v_W} +

3099: ne^{-\vert\bar{\omega}\vert a/v_W})} {(jn - km )} ,

3100: \eeq

3101: where 

3102: \bea

3103: j &=& (1+\frac{K_W}{K_C})e^{\vert\bar{\omega}\vert 

3104: d(\frac{1}{v_C}-\frac{1}{v_W})} + 

3105: \gamma_1 (1-\frac{K_W}{K_C})e^{-\vert\bar{\omega}\vert 

3106: d(\frac{1}{v_C}+\frac{1}{v_W})} \nonum \\

3107: k &=& (1-\frac{K_W}{K_C})e^{\vert\bar{\omega}\vert 

3108: d(\frac{1}{v_C}+\frac{1}{v_W})} +

3109: \gamma_1 (1+\frac{K_W}{K_C})e^{-\vert\bar{\omega}\vert

3110: d(\frac{1}{v_C}-\frac{1}{v_W})} \nonum \\

3111: m &=& (1-\frac{K_W}{K_C})e^{-\vert\bar{\omega}\vert

3112: (l+d)(\frac{1}{v_C}+\frac{1}{v_W})} +

3113: \gamma_1 (1+\frac{K_W}{K_C})e^{-\vert\bar{\omega}\vert

3114: (l+d)(\frac{1}{v_C}+\frac{1}{v_W})}e^{-2\vert\bar{\omega}\vert

3115: \frac{d}{v_C}} \nonum \\

3116: n &=& (1+\frac{K_W}{K_C})e^{-\vert\bar{\omega}\vert

3117: (l+d)(\frac{1}{v_C}-\frac{1}{v_W})} + 

3118: \gamma_1 (1-\frac{K_W}{K_C})e^{-\vert\bar{\omega}\vert

3119: (l+d)(\frac{1}{v_C}-\frac{1}{v_W})}e^{-2\vert\bar{\omega}\vert \frac{d}{v_C}}.

3120: \eea 

3121: Again, we will not give the results of taking the various limits 

3122: corresponding to the different frequency (or temperature) regimes 

3123: as these have already been quoted in the section on the conductance 

3124: of a quantum wire and quantum point contact.

3125: 

3126: \newpage

3127: 

3128: \begin{thebibliography}{99}

3129: 

3130: \bibitem{datta} S. Datta, {\it Electronic transport in mesoscopic systems}

3131: (Cambridge University Press, Cambridge, 1995); {\it Transport phenomenon in

3132: mesoscopic systems}, edited by H. Fukuyama and T. Ando (Springer Verlag,

3133: Berlin, 1992); Y. Imry, {\it Introduction to Mesoscopic Physics} (Oxford 

3134: University Press, New York, 1997).

3135: 

3136: \bibitem{wees} B. J. van Wees, H. van Houten, C. W. J. Beenakker, J. G.

3137: Williamson, L. P. Kouwenhoven, D. van der Marel and C. T. Foxon, Phys. Rev.

3138: Lett. {\bf 60}, 848 (1988); D. A. Wharam, T. J. Thornton, R. Newbury, M.

3139: Pepper, H. Ahmed, J. E. Frost, D. G. Hasko, D. C. Peacock, D. A. Ritchie and

3140: G. A. C. Jones, J. Phys. C {\bf 21}, L209 (1988).

3141: 

3142: \bibitem{tarucha} S. Tarucha, T. Honda and T. Saku, Sol. St. Comm. {\bf 94}, 

3143: 413 (1995). 

3144: 

3145: \bibitem{yacoby} A. Yacoby, H. L. Stormer, N. S. Wingreen, L. N. Pfeiffer, K.

3146: W. Baldwin and K. W. West, Phys. Rev. Lett. {\bf 77}, 4612 (1996).

3147: 

3148: \bibitem{thomas} K. J. Thomas, J. T. Nicholls, M. Y. Simmons, M. Pepper, D. R.

3149: Mace and D. A. Ritchie, Phys. Rev. Lett. {\bf 77}, 135 (1996).

3150: 

3151: \bibitem{facer} B. E. Kane, G. R. Facer, A. S. Dzurak, N. E. Lumpkin, R. G.

3152: Clark, L. N. Pfeiffer and K. W. West, App. Phys. Lett. {\bf 72}, 3506 (1998).

3153: 

3154: \bibitem{liang} C.-T. Liang, M. Pepper, M. Y. Simmons, C. G. Smith and D. A.

3155: Ritchie, Phys. Rev. B {\bf 61}, 9952 (2000).

3156: 

3157: \bibitem{liang2} C.-T. Liang, M. Y. Simmons, C. G. Smith, D. A. Ritchie and

3158: M. Pepper, App. Phys. Lett. {\bf 75}, 2975 (1999).

3159: 

3160: \bibitem{picciotto} R. de Picciotto, H. L. Stormer, A. Yacoby, L. N. Pfeiffer, 

3161: K. W. Baldwin and K. W. West, Phys. Rev. Lett. {\bf 85}, 1730 (2000).

3162: 

3163: \bibitem{topinka} M. A. Topinka, B. J. LeRoy, S. E. J. Shaw, E. J. Heller,

3164: R. M. Westervelt, K. D. Maranowski and A. C. Gossard, Science {bf 289}, 2323 

3165: (2000).

3166: 

3167: \bibitem{crook} R. Crook, C. G. Smith, M. Y. Simmons and D. A. Ritchie,

3168: cond-mat/9909017.

3169: 

3170: \bibitem{reilly} D. J. Reilly, G. R. Facer, A. S. Dzurak, B. E. Kane, R. G. 

3171: Clark, P. J. Stiles, J. L. O'Brien, N. E. Lumpkin, L. N. Pfeiffer and K. W. 

3172: West, cond-mat/0001174.

3173:  

3174: \bibitem{haldane} F. D. M. Haldane, Phys. Rev. Lett. {\bf 47}, 1840 (1981).

3175: 

3176: \bibitem{kane} C. L. Kane and M. P. A. Fisher, Phys. Rev. B {\bf 46}, 15233 

3177: (1992).

3178: 

3179: \bibitem{furusaki1} A. Furusaki and N. Nagaosa, Phys. Rev. B {\bf 47}, 4631 

3180: (1993).

3181: 

3182: \bibitem{safi} I. Safi and H. J. Schulz, Phys. Rev. B {\bf 52}, 17040 (1995).

3183: 

3184: \bibitem{maslov} D. L. Maslov and M. Stone, Phys. Rev. B {\bf 52}, 5539 

3185: (1995); V. V. Ponomarenko, {\it ibid.} {\bf 52}, R8666 (1995).

3186: 

3187: \bibitem{furusaki} A. Furusaki and N. Nagaosa, Phys. Rev. B {\bf 54}, 5239 

3188: (1996).

3189: 

3190: \bibitem{alekseev} A. Yu. Alekseev and V. V. Cheianov, Phys. Rev. B {\bf 57}, 

3191: 6834 (1998).

3192: 

3193: \bibitem{sumathi} For a review, see S. Rao and D. Sen, cond-mat/005492.

3194: 

3195: \bibitem{buttiker} M. B\"uttiker, Phys. Rev. B {\bf 41}, 7906 (1990).

3196: 

3197: \bibitem{fabrizio} M. Fabrizio and A. O. Gogolin, Phys. Rev. B {\bf 51}, 

3198: 17827 (1995).

3199: 

3200: \bibitem{chamon} C. de C. Chamon and E. Fradkin, Phys. Rev. B {\bf 55}, 4534 

3201: (1997); N. P. Sandler, C. C. Chamon and E. Fradkin, Phys. Rev. B {\bf 56}, 

3202: 2012 (1997).

3203: 

3204: \bibitem{safilong} I. Safi and H. J. Schulz, Phys. Rev. B {\bf 59}, 3040 

3205: (1999); I. Safi, Ph.D. thesis, Laboratoire de Physique des Solides, Orsay 

3206: (1996); I. Safi, Ann. Phys. (Paris) {\bf 22}, 463 (1997).

3207: 

3208: \bibitem{matveev} K. A. Matveev, Phys. Rev. B {\bf 51}, 1743 (1995).

3209: 

3210: \bibitem{yacimry} A. Yacoby and Y. Imry, Phys. Rev. B {\bf 41}, 5341 (1990).

3211: 

3212: \bibitem{maslovlong} D. L. Maslov, Phys. Rev. B {\bf 52}, R14386 (1995).

3213: 

3214: \bibitem{kimura} T. Kimura, K. Kuroki and H. Aoki, Phys. Rev. B {\bf 53}, 

3215: 9572 (1996).

3216:  

3217: \bibitem{tkachenko} O. A. Tkachenko, V. A. Tkachenko, D. G. Baksheyev, C.-T.

3218: Liang, M. Y. Simmons, C. G. Smith, D. A. Ritchie, G.-H. Kim and M. Pepper,

3219: cond-mat/0003021.

3220: 

3221: \bibitem{senz} V. Senz, T. Heinzel, T. Ihn, S. Lindemann, R. Held, K. Ensslin,

3222: W. Wegscheider and M. Bichler, cond-mat/0012205.

3223: 

3224: \end{thebibliography}

3225: 

3226: \newpage

3227: 

3228: \noindent {\bf Figure Captions}

3229: \vskip 1 true cm

3230: 

3231: \noindent {1.} Schematic diagram of the model showing the lead regions 

3232: (marked FL for Fermi liquid), the contact regions (C) of length $d$, and the 

3233: quantum wire (QW) of length $l$. The interaction parameters in these three 

3234: regions are denoted by $K_L$, $K_C$ and $K_W$ respectively.

3235: 

3236: \noindent {2.} Diagram showing the Fermi energy $E_F$ (with a thermal spread 

3237: of $k_B T$) in relation to the sub-bands within the quantum wire. The 

3238: conductance will lie on a plateau if the energy levels are as shown in (a),

3239: while the conductance will be at a step between one plateau and the next if 

3240: the energy levels are as shown in (b).

3241: 

3242: \noindent {3.} A plot showing the curve fitted to the conductance corrections 

3243: $\delta g$ versus temperature $T$ data obtained from the inset of Fig. 3 of 

3244: Yacoby {\it et al} \cite{yacoby}. The expression for the curve and the value 

3245: of the correlation coefficient are given in the text.

3246: 

3247: \newpage

3248: 

3249: \begin{figure}

3250: \begin{center}

3251: \epsfig{figure=fig1.ps,bbllx=50,bblly=0,bburx=550,bbury=800,height=21cm}

3252: \end{center}

3253: \vspace*{-3 cm}

3254: \centerline{Fig. 1}

3255: \end{figure}

3256: 

3257: \begin{figure}

3258: \begin{center}

3259: %\epsfig{figure=fig2.ps,angle=-90,bbllx=50,bblly=0,

3260: %bburx=500,bbury=800,height=15cm}

3261: \epsfig{figure=fig2.ps,angle=-90,width=16cm}

3262: \end{center}

3263: \vspace*{2 cm}

3264: \centerline{Fig. 2}

3265: \end{figure}

3266: 

3267: \begin{figure}

3268: \begin{center}

3269: \epsfig{figure=fig3.ps,bbllx=50,bblly=-200,bburx=550,bbury=600,height=21cm}

3270: \end{center}

3271: \vspace*{-5 cm}

3272: \centerline{Fig. 3}

3273: \end{figure}

3274: 

3275: \end{document}

3276: 

3277: