Multiband Effects in Fe- -pnictide Superconductors pnictide - PowerPoint PPT Presentation

Multiband Effects in Fe- -pnictide Superconductors pnictide Superconductors Multiband Effects in Fe Zlatko Tesanovic Johns Hopkins University E-mail: zbt@pha.jhu.edu Web: http://www.pha.jhu.edu/~zbt V. Cvetkovic and ZT , arXiv/0804.4678 V. Cvetkovic and ZT , arXiv/0808.3742 T. Y. Chen et al., Nature 453 , 1224 (2008) Iron Pnictides – Semimetals turned Superconductors o Minimal Model of FeAs planes – Different from CuO 2 !! o Superconducting Gap in FeAs (PCAR, ARPES, mw, STM) o Multiband Magnetism and Superconductivity in FeAs o Summer Blockbuster of 2008 Summer Blockbuster of 2008 9/26/2008

Pnictides: Semiconductors � Semimetals � Superconductors Pnictides – elements from Group V of Periodic Table: nitrogen, phosphorus, arsenic, antimony and bismuth III-V Semiconductors – formed by elements from Groups III and V: aluminium phosphide, aluminium arsenide, aluminium antimonide, gallium phosphide, gallium arsenide, gallium antimonide, indium phosphide, indium arsenide and indium antimonide plus numerous ternary and quaternary semiconductors.

Fe-pnictides: Semimetals � Superconductors Iron Pnictides Spring 2008 Greatest web-induced frenzy in history of condensed matter physics: 17 papers on arXiv in a single July day. Comparable to the latest superstring “revolution” (Bagger-Lambert)

Fe-pnictides: Semimetals � Superconductors May 2006 Hideo Hosono, TITech

Cu-oxides versus Fe-pnictides As Fe O RE 1111 122 122 However, there are also many differences! This may add up to a new and interesting physics

Key Difference: 9 versus 6 d-electrons Even (FeAs) vs odd (CuO 2 ) number of d-electrons Nearly half-filled (FeAs) vs nearly-filled (CuO 2 ) d-shell In CuO 2 one d-hole in a half-filled single band � In FeAs large and even number of d-holes/several d-like bands �

Fe-pnictides: Semimetals � Superconductors In contrast to CuO 2 , all d- bands in FeAs are either nearly empty (electrons) or nearly full (holes) and far from being half-filled. FeAs are less � correlated than CuO 2

Phase diagram of Fe-pnictides C. de la Cruz, et al ., Nature 453 , 899 (2008) Like CuO 2 , phase diagram FeAs has SDW (AF) in proximity to the SC state. H. Chen, et al ., arXiv/0807.3950 SC coexists with SDW (AF) in 122 compounds �

Superconductivity in Fe K. Shimizu et al . Nature 412 , 316-318 (2001).

Minimal Model of FeAs layers I V. Cvetkovic and ZT, arXiv/0804.4678 “Puckering” of FeAs planes is essential: i)All d-orbitals are near E F ii)Large overlap with As p-orbitals below E F � enhanced itinerancy of d electrons defeats Hund’s rule and large local moment

Hund’s Rule Defeated V. Cvetkovic and ZT, arXiv/0804.4678 Hund’s rule rules for Mn 2+ : all five d-electrons line up to minimize Coulomb repulsion � S = 5/2 “Puckering” of FeAs planes is essential: i)All d-orbitals are near E F ii)Large overlap with As p-orbitals below E F � enhanced itinerancy of d electrons defeats Hund’s rule and large local moment Haule, Shim and Kotliar, PRL 100 , 226402 (2008)

Minimal Model of FeAs layers II V. Cvetkovic and ZT, arXiv/0804.4678 Important: Near E F e and h bands contain significant admixture of all five Wannier d-orbitals, d xz and d yz of odd parity in FeAs plane) and the remaining three d-orbitals of even parity in FeAs plane �

Valley Density-Wave (VDW) in Fe-pnictides V. Cvetkovic and ZT, arXiv/0804.4678 V. Cvetkovic and ZT , arXiv/0808.3742 Outcome: CDW (structural) and SDW (AF) orders at q = M

Fictitious “Superconductor” � VDW in Fe-pnictides V. Cvetkovic and ZT , arXiv/0808.3742 V. Cvetkovic and ZT, arXiv/0804.4678 ? What about real superconductivity ?

Coastline of the Fermi Sea + Fermi sea Fermi sea - + New REOFeAs SC T c ∼ 55K - - Fermi sea +

What can Δ tell us about superconducting state ? virtual phonons electrons Cooper pair size = coherence length ξ Standard BCS theory works well in materials like Nb, Sn or Hg. In Pb and more complex systems (Va 3 Sn) one needs “strong coupling” theory (2 Δ /T c ∼ 4-6 )

What can Δ tell us about superconducting state ? + - - Fermi sea + Our results for FeAs appear inconsistent with these features

Andreev Spectroscopy Andreev Spectroscopy Experimental setup Gold tip in contact with FeAs SC Gap value from Andreev peaks 2 Δ ≈ 13.4 meV 2 Δ /k B T C = 3.68 (BCS) Extra features beyond gap (contact specific) Slanted background [Always G(-V) > G(V) ] 18

Δ in FeAs superconductors I T. Y. Chen et al., Nature 453 , 1224 (2008) Conclusions: Nodeless superconducting gap and no pseudogap behavior. Very different from high temperature cuprate superconductors !!

Δ in FeAs superconductors II V. Cvetkovic and ZT, arXiv/0804.4678 Only a single superconducting gap – sign/phase could be different for holes and electrons. Conclusions: Conventional phonon-mechanism is unlikely but so is Mott limit-induced repulsion of the cuprate d-wave kind. We have something new !!

Emergence of Zero Bias Anomaly (ZBA) Emergence of Zero Bias Anomaly (ZBA) Small contact Contact size • Emerge systematically • ZBA obscures gap Large • ZBA due to SC contact • H field has small effect at 4.5K

ZBA in s- -wave Nb wave Nb ZBA in s G(V) -1 G(V) P. Xiong, G. Xiao, R. B. Laibowitz, Phys. Rev. Lett. 71 , 1907 (1993)

Nb tip on Cu thin film (Chen et al et al ) ) Nb tip on Cu thin film (Chen

Nb tip on Cu thin film (Chen et al et al ) ) Nb tip on Cu thin film (Chen 7.0 10.7 Ω 6.5 d 1 d 2 9.7 Ω 6.0 d 1 d 2 5.5 6.6 Ω 5.0 d 3 d 1 d 2 4.5 4.4 Ω d 1 d 2 4.0 d 3 dI/dV 3.1 Ω 3.5 d 4 d 2 d 3 3.0 2.3 Ω , 1.27 Ω d 4 2.5 d 2 d 5 d 3 2.0 3.5 Ω , 1.19 Ω d 4 d 5 d 2 1.5 d 3 4.4 Ω , 0.62 Ω 1.0 d 4 d 2 0.5 -18 -15 -12 -9 -6 -3 0 3 6 9 12 15 18 V (mV)

Minimal Model of FeAs layers III V. Cvetkovic and ZT, arXiv/0804.4678 h 2 h 1 e 1 FeAs are different from CuO 2 Charge carriers are more itinerant and less localized on atomic sites. Multiband description is necessary, unlike an effective single band model of cuprates

Interactions in FeAs I V. Cvetkovic and ZT, arXiv/0804.4678

Interactions in FeAs II V. Cvetkovic and ZT, arXiv/0804.4678 V. Cvetkovic and ZT , arXiv/0808.3742 h 1 h 2 e 1 e-h Typically, we find W s is dominant � These “Josephson” terms do not appear in fictitious Valley density-wave(s) (VDW) in FeAs superconductor analogy � Could they cause real SC ?

Valley Density-Wave (VDW) and SC in FeAs V. Cvetkovic and ZT, arXiv/0808.3742 V. Stanev, J. Kang, ZT, arXiv/0809.0014 Outcome: combined SDW (AF) and CDW/structural orders at q = M “Josephson” terms in k- VDW �� SC space (c + c + dd) play key role in SC Near VDW transition VDW fluctuations enhance interband “Josephson” repulsion. SC state with ∆ (holes) and - ∆ (electrons). �

THE END THE END

Valley Density-Wave (VDW) in Fe-pnictides V. Cvetkovic and ZT, arXiv/0804.4678 V. Cvetkovic and ZT , arXiv/0808.3742 Outcome: CDW (structural) and SDW (AF) orders at q = M

Valley Density-Wave (VDW) in Fe-pnictides V. Cvetkovic and ZT, arXiv/0804.4678 V. Cvetkovic and ZT , arXiv/0808.3742 Outcome: CDW (structural) and SDW (AF) orders at q = M

BCS-like gap from BTK analysis At 4.52 K 2 Δ = 13.34 ± 0.3 meV T C = 42 K 2 Δ /k B T C = 3.68 closer to 3.53 (BCS s -wave) than 4.28 (BCS d -wave)