Abstract
The correct definition of the conductance of finite systems implies a connection to the system of the massive ideal leads. Influence of the latter on the properties of the system appears to be rather essential and is studied below on the simplest example of the 1D case. In the log-normal regime, this influence is reduced to the change of the absolute scale of conductance, but generally changes the whole distribution function. Under the change of the system length L, its resistance may undergo the periodic or aperiodic oscillations. Variation of the Fermi level induces qualitative changes in the conductance distribution, resembling the smoothed Anderson transition.
Similar content being viewed by others
Notes
It should be stressed, that the limit L → 0 is indeed formal, since the results (7) and (8) are restricted by the condition L\( \gtrsim \) 1/κ.
In particular, the analysis of fourth moments for matrices with the complex elements looks rather hopeless, since it demands diagonalization of the matrix of large size.
We have listed all possible combinations. Indeed, a change of the δ sign in (43), (47) leads to the analogous matrices, which can be transformed to the initial form, if the components of columns are renumbered in the inverse order; so the odd powers of δ do not appear. Since we consider the limit δ ~ \({{\epsilon }^{2}}\) → 0 (see Eq. (45)), then only combinations δ2n\({{\epsilon }^{{2m}}}\) with 4n + 2m ≤ 6 for (43) and 4n + 2m ≤ 10 for (47) are possible; among them only combinations with n ≥ m do not have singularities for δ → 0.
For this, strictly speaking, one should average (32) over the n variations of order 1/δ.
Let us remind that the Fourier transform of P(ρ) gives the characteristic function F(t) = 〈eiρt〉, which is the generating function of moments, F(t) = \(\sum\nolimits_{n = 0}^\infty {{{{(it)}}^{n}}} \)〈ρn〉/n!. If all moments of the distribution are known, then one can construct F(t), while P(ρ) is given by the inverse Fourier transform.
Comparison of (44) and (48) shows that the periods of oscillations for 〈ρ〉 and 〈ρ2〉 differ by a factor 211/3. As clear from Section 4 (see Footnote 3), the right hand side of the equation for x may contain only combinations δ2n\({{\epsilon }^{{2m}}}\) with n ≥ m, of which only δ2n\({{\epsilon }^{{2n}}}\) ~ W2n remain finite in the δ → 0 limit. Since for x ~ δ ~ \({{\epsilon }^{2}}\) all terms of the equation have the same order of magnitude, the exponent of growth x for 〈ρn〉 at δ = 0 satisfies the equation x2n+ 1 = c1δ2\({{\epsilon }^{2}}\)x2n– 2 + c2δ4\({{\epsilon }^{4}}\)x2n– 5 + …, whose nontrivial roots are of order (δ2\({{\epsilon }^{2}}\))1/3 independently of n.
In this case the parameter γ ~ 1 and does not have the essential evolution.
REFERENCES
P. W. Anderson, D. J. Thouless, E. Abrahams, and D. S. Fisher, Phys. Rev. B 22, 3519 (1980).
R. Landauer, IBM J. Res. Dev. 1, 223 (1957);
Philos. Mag. 21, 863 (1970);
E. N. Economou and C. M. Soukoulis, Phys. Rev. Lett. 46, 618 (1981).
D. S. Fisher and P. A. Lee, Phys. Rev. B 23, 6851 (1981).
D. C. Langreth and E. Abrahams, Phys. Rev. B 24, 2978 (1981).
P. W. Anderson, Phys. Rev. B 23, 4828 (1981);
H. L. Engquist and P. W. Anderson, Phys. Rev. B 24, 1151 (1981).
D. J. Thouless, Phys. Rev. Lett. 47, 972 (1981).
M. Buttiker, Phys. Rev. Lett. 57, 1761 (1986);
M. Buttiker, Y. Imry, R. Landauer, and S. Pinhas, Phys. Rev. B 31, 6207 (1985).
M. Ya. Azbel, J. Phys. C 14, L225 (1981).
A. D. Stone and A. Szafer, IBM J. Res. Dev. 32, 384 (1988).
I. M. Suslov, J. Exp. Theor. Phys. 115, 897 (2012).
I. M. Suslov, J. Exp. Theor. Phys. 124, 763 (2017).
V. I. Melnikov, Sov. Phys. Solid State 23, 444 (1981).
A. A. Abrikosov, Solid State Commun. 37, 997 (1981).
N. Kumar, Phys. Rev. B 31, 5513 (1985).
B. Shapiro, Phys. Rev. B 34, 4394 (1986).
P. Mello, Phys. Rev. B 35, 1082 (1987).
B. Shapiro, Philos. Mag. 56, 1031 (1987).
I. M. Lifshitz, S. A. Gredeskul, and L. A. Pastur, Introduction to the Theory of Disordered Systems (Nauka, Moscow, 1982) [in Russian].
X. Chang, X. Ma, M. Yepez, A. Z. Genack, and P. A. Mello, Phys. Rev. B 96, 180203 (2017).
L. I. Deych, D. Zaslavsky, and A. A. Lisyansky, Phys. Rev. Lett. 81, 5390 (1998).
L. I. Deych, A. A. Lisyansky, and B. L. Altshuler, Phys. Rev. Lett. 84, 2678 (2000);
Phys. Rev. B 64, 224202 (2001).
L. I. Deych, M. V. Erementchouk, and A. A. Lisyansky, Phys. Rev. Lett. 90, 126601 (2001).
L. D. Landau and E. M. Lifshitz, Course of Theoretical Physics, Vol. 3: Quantum Mechanics: Non-relativistic Theory (Nauka, Moscow, 1989; Pergamon, New York, 1977).
B. L. Altshuler, JETP Lett. 41, 648 (1985);
B. L. Altshuler and D. E. Khmelnitskii, JETP Lett. 42, 359 (1985).
P. A. Lee and A. D. Stone, Phys. Rev. Lett. 55, 1622 (1985);
P. A. Lee, A. D. Stone, and H. Fukuyama, Phys. Rev. B 35, 1039 (1987).
D. Braun et al., Phys. Rev. B 64, 155107 (2001).
I. Travenec, Phys. Rev. B 69, 033094 (2004).
M. Ruhlender, P. Markos, and C. M. Soukoulis, Phys. Rev. B 64, 172202 (2001).
A. O. Gel’fand, Calculus of Finite Differences (Nauka, Moscow, 1967) [in Russian].
I. M. Suslov, J. Exp. Theor. Phys. 127, 131 (2018).
M. Kappus and F. Wegner, Z. Phys. B 45, 15 (1981).
V. E. Kravtsov and V. I. Yudson, Ann. Phys. (N.Y.) 326, 1672 (2011).
S. Washburn and R. A. Webb, Adv. Phys. 35, 375 (1986).
D. Mailly and M. Sanquer, J. Phys. (France) I 2, 357 (1992).
E. Abrahams, P. W. Anderson, D. C. Licciardello, and T. V. Ramakrishman, Phys. Rev. Lett. 42, 673 (1979).
A. Cohen, Y. Roth, and B. Shapiro, Phys. Rev. B 38, 12125 (1988).
I. M. Suslov, Philos. Mag. 99, 247 (2019).
Author information
Authors and Affiliations
Corresponding author
Appendices
EVOLUTION OF MOMENTS FOR THE COORDINATE TRANSFER MATRIX
For the coordinate transfer matrix the following evolution equations are valid
where zn = \(\tau _{{12}}^{{(n - 1)}}\), yn = \(\tau _{{11}}^{{(n - 1)}}\) with the initial conditions z1 = 0, y1 = 1, or zn = \(\tau _{{22}}^{{(n - 1)}}\), yn = \(\tau _{{21}}^{{(n - 1)}}\) with the initial conditions z1 = 1, y1 = 0. For the second moments one has
and suggesting their exponential behavior λn with λ = 1 + x, obtains the equation for x
Setting E2 – 4 = 4δ2, W2 = 4\({{\epsilon }^{2}}\)δ2 and taking the limit δ → 0, \(\epsilon \) → 0, δ/\({{\epsilon }^{2}}\) = const, one can verify that (A.3) coincides with (44).
Analogously, for the fourth moments one has a system of equations
Suggesting that the moments behave as λn with λ = 1 + x, we have the equation for x
where terms of the higher order in W2 are omitted. Setting E2 – 4 = 4δ2, W2 = 4\({{\epsilon }^{2}}\)δ2 and retaining the main terms at the indicated limiting transition, we come to Eq. (48).
THE ASYMPTOTIC FORMS OF THE INTEGRAL (70)
Using the evenness of the integrand of (70) in variables \(\tilde {\varphi }\) = φ – π/2, \(\tilde {\theta }\) = θ – π/2 and setting x = sinφ, y = sinθ, we have
Configuration of saddle points is essentially different for ρ < ρc and ρ > ρc, where ρc = \(\Delta _{1}^{2}\). In the first case the maximum of the exponent is reached at x = xc, y = –1 or x = –xc, y = 1, where xc = Δ2\(\sqrt \rho \)/Δ1\(\sqrt {1 + \rho } \), while in the second case at x = 1, y = –1 or x = –1, y = 1. Let δx and δy are deviations of x and y from the extremum point. For ρ < ρc we retain in S(x, y) the quadratic term in δx and the linear term in δy, setting x = xc, 1 – y2 = 2δy in the pre-exponential; then
which is the limiting form of the distribution for t → 0.
For ρ > ρc we have
where Sc, A, B are defined in Eq. (73) and C = \(\Delta _{1}^{2}\)(1 + ρ). For large ρ the linear terms in δx, δy are sufficient in S(x, y) and in radicals; then
which is the last asymptotics in (72), (74)–(77).
The results (B.2), (B.4) are not applicable for ρ close to ρc, since the coefficient A turns to zero at ρ = ρc. Setting δ\(\tilde {x}\) = δx – xcδy and omitting the last term in (B.3), we have
For A2 ≪ Ct fluctuations of δ\(\tilde {x}\) are determined by the quadratic term, and Aδ\(\tilde {x}\) can be omitted; then in the small vicinity of ρc
so divergency at ρ → ρc is eliminated and the third asymptotics in (72) and (75) is recovered. Under the indicated condition, this result remains valid for ρ < ρc, when A is negative, but small in modulus. In fact, Eq. (B.6) takes place in the cases t ≪ \(\Delta _{1}^{2}\) ≪ 1 and t ≪ 1 ≪ \(\Delta _{1}^{2}\), when the fluctuation δy ~ t/ρ is small in comparison with δ\(\tilde {x}\) ~ \(\sqrt {t{\text{/}}C} \), so δ\(\tilde {x}\) ≈ δx and one can set 1 – x2 = 2δ\(\tilde {x}\), 1 – y2 = 2δy in the pre-exponential. The inverse situation is realized in the case \(\Delta _{1}^{2}\) ≫ 1, t ≫ 1, when δy ≫ δ\(\tilde {x}\) and 1 – x2 ≈ 2δx = 2δ\(\tilde {x}\) + 2xcδy ≈ 2δy in the preexponential; then we come to the result
determining the penultimate asymptotics in (76), (77). The condition A2 ≪ Ct corresponds to (ρ – ρc)2 ≪ \(\Delta _{1}^{2}\)ρt, which for t ≫ 1 is reduced to the given in (B.7). If (B.6) is valid in the small vicinity of ρc, the wide range of validity arises for (B.7).
For small ρ the saddle point approximation is not applicable, and the initial form of (70) is convenient. If t ≪ \(\Delta _{1}^{2}\), then saddle point integration over φ is still possible and leads to expression
For \(\Delta _{1}^{2}\) ≪ 1 one has Δ2 ≈ 1; if ρ ≪ t, then only \(\Delta _{1}^{2}\) can be retained in the denominator, while integration over θ can be produced by expansion of the exponent
which is the first result in (72), (75), (76). For ρ \( \gtrsim \)t one returns to the saddle point results (B.2), (B.4), (B.5).
In the case \(\Delta _{1}^{2}\) ≪ 1 and t ≫ \(\Delta _{1}^{2}\) the region of small ρ is determined by the condition ρ(1 + ρ) ≪ t2/\(\Delta _{1}^{2}\) and the integral (70) is calculated by expansion over S(φ, θ)/t till the second order, which gives the first result in (74), (77); for the opposite inequality we have A ≈ B ≈ 2Δ1Δ2\(\sqrt {\rho (1 + \rho )} \) ≫ t and A2 ≫ Ct, which is sufficient for validity of (B.4).
In the case \(\Delta _{1}^{2}\) ≫ 1 and t ≪ \(\Delta _{1}^{2}\)Eq. (B.8) remains valid, but its analysis is more complicated. For t ≪ 1 we have the familiar situation: in the interval ρ \( \lesssim \)t one can retain \(\Delta _{1}^{2}\) in the denominator and obtain (B.9), while in the interval ρ \( \gtrsim \)t the saddle point results (B.2), (B.4), (B.5) are valid. For 1 ≪ t ≪ \(\Delta _{1}^{2}\) the result (B.9) is valid only for ρ \( \lesssim \) 1. In the interval 1 \( \lesssim \) ρ \( \lesssim \)t the term \(\Delta _{2}^{2}\)ρsin2θ is dominated in the denominator of (B.8), while the quantity \(\Delta _{1}^{2}\) – ρ is necessary only for cutoff of the logarithmic divergency:
In the interval t\( \lesssim \) ρ \( \lesssim \) ρc the exponent restricts integration in (B.8) by values θ2\( \lesssim \)t/ρ, so \(\Delta _{1}^{2}\)ρ in (B.10) is changed by \(\Delta _{1}^{2}\)t. In fact, both results are actual only for ρ ≪ ρc, since in the interval ρ \( \gtrsim \) ρc/t divergency at ρ → ρc is eliminated due to nonlinear terms in (B.3) and the result (B.7) is valid; so we have ln ρ for 1 \( \lesssim \) ρ \( \lesssim \)t and ln t for t\( \lesssim \) ρ \( \lesssim \) ρc/t, as is reflected in (76).
In the case \(\Delta _{1}^{2}\) ≫ 1 and t ≫ \(\Delta _{1}^{2}\), expansion over S(φ, θ)/t is possible in the interval ρ \( \lesssim \)t/\(\Delta _{1}^{2}\) and leads to the first result (77). In the interval ρ \( \gtrsim \)t/\(\Delta _{1}^{2}\) expression (B.8) is valid, where |\(\Delta _{1}^{2}\) – ρ| ≪ t, \(\Delta _{2}^{2}\) ≫ ρ≫ t and the latter term is dominant in the denominator; the logarithmic divergency is removed due to restriction (\(\Delta _{2}^{2}\)ρ/t)sin2θ \( \gtrsim \) 1, which is necessary for the saddle point integration over φ and validity of (B.8). If ρ \( \lesssim \)t, then the exponent in (B.8) is not essential and the second result (77) holds. If ρ \( \gtrsim \)t, then we have the saddle point situation and validity of (B.7) and (B.4).
Rights and permissions
About this article
Cite this article
Suslov, I.M. Conductance Distribution in 1D Systems: Dependence on the Fermi Level and the Ideal Leads. J. Exp. Theor. Phys. 129, 877–895 (2019). https://doi.org/10.1134/S1063776119090139
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1134/S1063776119090139