1996
DOI: 10.1103/physrevlett.77.3909
|View full text |Cite
|
Sign up to set email alerts
|

Theory of Thermal Conductivity inYBa2Cu3O7

Abstract: We calculate the electronic thermal conductivity in a d-wave superconductor, including both the effect of impurity scattering and inelastic scattering by antiferromagnetic spin fluctuations. We argue that phonons dominate heat transport near T c , but that electrons are responsible for most of the peak observed in clean samples, The peak position is predicted to vary nonmonotonically with disorder, in good agreement with experiments on YBa 2 ͑Cu 12x Zn x ͒ 3 O 72d . [S0031-9007(96)01535-9] PACS numbers: 74.72.… Show more

Help me understand this report
View preprint versions

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1

Citation Types

9
70
0

Year Published

1997
1997
2015
2015

Publication Types

Select...
8
1
1

Relationship

2
8

Authors

Journals

citations
Cited by 86 publications
(79 citation statements)
references
References 24 publications
9
70
0
Order By: Relevance
“…2b), we show the low-T thermal conductivity for the case of n i = 2.5% (15%) scatterers with V imp /t = 10 (V imp /t = 3) averaged over 20 different random impurity configurations (which is enough for configurational convergence). For U = 0 the result in 2a agrees well with those obtained previously for the dilute impurity limit: in the non-self-consistent (NSC) calculation (with homogeneous ∆=0.4), κ(T )/T = κ 00 + αT 2 [31], whereas in the self-consistent (SC) case the spatial inhomogeneity of ∆ ij causes a reduction of α and leads to a more linear T dependence [9] of κ/T . The case with V imp /t = 3.0 in 2b is an example of one of the weak- est disorder potentials we are able to study since further reduced V imp leads to mean-free-path larger that our system.…”
supporting
confidence: 80%
“…2b), we show the low-T thermal conductivity for the case of n i = 2.5% (15%) scatterers with V imp /t = 10 (V imp /t = 3) averaged over 20 different random impurity configurations (which is enough for configurational convergence). For U = 0 the result in 2a agrees well with those obtained previously for the dilute impurity limit: in the non-self-consistent (NSC) calculation (with homogeneous ∆=0.4), κ(T )/T = κ 00 + αT 2 [31], whereas in the self-consistent (SC) case the spatial inhomogeneity of ∆ ij causes a reduction of α and leads to a more linear T dependence [9] of κ/T . The case with V imp /t = 3.0 in 2b is an example of one of the weak- est disorder potentials we are able to study since further reduced V imp leads to mean-free-path larger that our system.…”
supporting
confidence: 80%
“…[6] and suggested to correspond to Cu vacancies in the CuO 2 plane. While measured resonance energies on native defects and Zn are roughly consistent with fits to bulk transport measurements based on naive point-like potential scattering models 12 …”
Section: Introductionsupporting
confidence: 55%
“…The magnetic field is applied perpendicular to the ab plane and the H-dependence of κ is measured with the field-cooled procedure [9] to avoid complications that are associated with the vortex-pinning-related hysteresis [12]. Figure 2 displays the temperature dependences of ρ and κ for the a and b directions; the resistivity shows only a small, if any, in-plane anisotropy, while the thermal conductivity demonstrates a non-trivial anisotropy that cannot be attributed to an error in the geometrical factors; namely, the peak height at ∼70 K is smaller along the a axis and yet κ a becomes larger than κ b below ∼30 K. This anisotropy in κ(T ) suggests that the number of thermally-excited QPs is larger for the nodes along the a axis, because the QP population near the nodes is more directly reflected in κ at lower temperatures [13]. (The anisotropy in κ above T c is most likely due to phonons; it has been estimated [9] that phonons are responsible for roughly 2/3 of the total κ at T c .)…”
mentioning
confidence: 92%