X-ray vision of High Temperature Superconductivity Giacomo Ghiringhelli

Transcription

1 Firma convenzione Politecnico di Milano e Veneranda Fabbrica del Duomo di Milano X-ray vision of High Temperature Superconductivity Giacomo Ghiringhelli Dipartimento di Fisica Politecnico di Milano Italy Colloquium di Dottorato Aula Magna Rettorato Mercoledì 27 maggio 2015 Università di Pavia, Dipartimento di Fisica - 11 Maggio 2017

2

3 dvanced -Ray mission pectroscopy Soft x-rays, spectra and Lucio : ID12B : ID : ID32 INFM A X E S Lucio Braicovich : SLS SAXES & Swiss Light Source Politecnico di Milano

4 Very low T Superconductivity Discovered in 1911 in mercury Electric resistivity is ZERO below a critical temperature T c Resistance T (K)

5 Perfect diamagnetism Not only a perfect conductor, but also a perfect anti-magnet : Meissner effect A superconductor repels magnetic fields

6 How can we use SC? Transport and storage of electric power: no waste of energy Limitations Low T Critical current High magnetic fields: no need to dissipate heat from the coils of electromagnets Critical current Critical field Low T

7 SC technology today Despite the difficulty due to cryogenics, superconducting wires have been developed for special applications and are currently used in commercial medical devices and advanced scientific instrumentation. LHe Nb alloys MRI High field magnets Particle accelerators SQUID magnetometers

8 Low temperatures For 75 years the search for materials with higher T c has been quite frustrating Liquid He

9 High Tc superconductors (HTS) Discovery: Jan Publication: April 1986 Bednorz & Müller Nobel 1987

10 A new cult is born during the Woodstock of Physics (1987) American Physical Society, March Meeting, New York, 1987 The arxiv was invented to keep track of discoveries on HTS: normal editorial procedures were too slow to guarantee priority claims

11 YBCO: T c above 77 K Liquid N 2 Liquid He

12 Where are we with T c 30 years later? Cu based Fe based

13 New hope from Cu based HTS Higher T c : above Liquid N 2, cryogenics is cheaper and simpler Higher critical field: stronger electromagnets (potentially up to 40T) Higher critical current, provided we learn how to make good (multifilament) wires Multifilament: to reduce losses due to hysteresis, which is related to vortex motion, that give also energy losses

14 Future applications with HTS? Source: Carmine Senatore, Université de Genève - Lectures on Superconductivity and applications

15 Cuprates: structure CuO 2 planes O Cu

16 AntiFerroMagnetic order in CuO 2 planes O Cu 2+, 3d 9 O Cu O O AFM order in the CuO 2 planes INSULATOR

17 Doping of planes: holes O Cu 2+, 3d 9 O Cu O O The perfect (long range) AFM order is broken Superconductivity becomes possible

18 Phase diagram electrons holes

19 Extreme Complexity Nature Physics, 8, 864 (2012) All experimental techniques have been used to unveil the secrets of cuprates, but in most cases results have added complexity rather than clarity Nature Physics 12, 47 (2016)

20 What is special in HTS? The common property is in the CuO 2 planes, ie the system is considered quasi 2D. Thus all properties are referred to the squared 2D reciprocal lattice. Quasi 2D STRUCTURE RECIPROCAL SPACE For example: The electronic band structure as determined by Angle Resolved PhotoElectron Spectroscopy (ARPES) nuclear BZ Γ (0,0) M (½, ½) (π, π) X (½,0) (π,0) magnetic BZ

21 The Superconducting gap By ARPES we can determine the Fermi surface shape and the SC gap. The gap is not constant long the Fermi surface, but has x 2 -y 2 symmetry But ARPES cannot tell us what is the glue of Cooper pairs and Pseudo-Gap

22 Enter X-ray spectroscopy! XAS: ground state orbital symmetry XMCD: weak ferromagnetism and proximity effect Spin-Resolved XPS: spin character of doped sites RIXS: crystal field splitting RIXS: spin excitations (magnons) R(I)XS: charge density waves IXS: phonon dispersion RIXS: electron-phonon coupling

23 XAS: 3d hole symmetry Linear dichroism of Cu L 2,3 XAS was used to probe the symmetry of the empty Cu 3d orbital Cu hν θ E Result: the hole in Cu 2+ has 100% x 2 -y 2 symmetry

24 XMCD: weak ferromagnetism and proximity effect The spin moments are not perfectly in-plane: they are tilted by few degrees and a strong external field can orient their out-of-plane component and we can detect it with XMCD A ferromagnetic oxide can act similarly (proximity effect) LSMO LSCO G.M. De Luca, M. Salluzzo, GG et al, PRB 82, (2010); Nat. Comm (2014)

25 Spin-Resolved XPS: Zhang-Rice singlets How are the spins aligned when two holes sit at the same Cu site? Theory predict antiparallel configuration (Zhang-Rice singlet) A smart combination of resonant photoemission made with circularly polarized photons and by detecting the spin polarization of photoelectrons has confirmed it. O N.B. Brookes, GG et al, PRL (2001); PRL 115, (2015)

26 Introduction to Resonant X-ray Scattering X-ray Absorption Spectroscopy XAS ELS Energy Loss Spectroscopy RXS XRD X-Ray Diffraction

27 From XRD to X-ray Scattering θ crystal 2θ X-Ray Diffraction Real space Bragg law k' = k λ = λ' k' = k λ = λ' X-Ray Scattering Reciprocal lattice q=g q=g λ λ' (0,2) k k' (0,1) (1,1) q = k' - k (0,0) (1,0)

28 ELS: from Raman to Inelastic X-ray Scattering Energy Loss Spectroscopy Raman light scattering ω k k' ω' Inelastic X-ray Scattering k 0, q 0, Ω=ω ω' X-ray photons ω ω' Ω=ω ω' q = k' - k ω ω' Ω Ω

29 RIXS Resonant Inelastic X-ray Scattering Resonant X-ray Absorption RIXS Inelastic X-ray Scattering ω, λ, ε k q k' ω', λ', ε' X-Ray Scattering Ω=ω ω' q = k' k ε, ε ': polarization

30 E v Why soft x-rays for 3d transition E metals 3dTM Spin-Orbit splitting E F 4sp 3d Mn L 2,3 XAS hν in 3p ω La 0.7 Sr 0.3 MnO 3 2p MnO 1s photon energy (ev)

31 3dTM Core level binding energies and edges C O Si Sc Fe Zn Y Mo Cd Ce Gd Lu Au Th 3dTM 4dTM RE Actinides Binding energy (ev) Hard X-Rays Soft X-Rays UV 10 1s 2p 3/2 3p 3/2 3d 5/2 4p 3/2 4d 5/ Atomic number Z

32 E Cu L 3 RIXS dd Spin Phonons E F 3d hν out - hν in e-h E loss 0 Sr 2 CuO 2 Cl 2 hν in (930 ev for Cu L 3 ) hν out π σ q=+0.37 rlu xy 2p 3/2 2M Elastic 1M Energy loss (ev)

33 10 Cu L 3 RIXS: Experimental conditions 0,0 Wavevector of particles used in inelastic scattering Thermal neutrons 0.5,0.5 π,π Cu L 3 RXS 0.5,0 π,0 Cu L 3 resonance: E 0 = 930 ev q max = 0.86 Ang -1 confined inside a region around Γ 2p core hole: spin-orbit interaction E resolution: mev q resolution: rlu ½ - 1 hour per spectrum k (Ang -1 ) E-3 Neutrons 1st Brillouin zone boundary Photons 1m 10m 100m k 10k 100k energy (ev) 3dTM M edges 3dTM L edges 3dTM K edges, 5dTM L edges E L 3 RIXS INS q

34 RIXS: crystal field How much are the 3d orbitals separated in energy in the cuprates (crystal field splitting)? What is the relation to the coordination and atomic distances? RIXS has determined these values. d states x 2 -y 2, z 2 10Dq xy, yz,zx e g t 2g x 2 -y 2 z 2 10Dq xy yz,zx b 1 a 1 b 2 e g Spherical O 3 Cubic O h Tetragonal D 4h M. Moretti Sala, L. Braicovich, GG et al, New J. Phys. 13, (2011)

35 RIXS: spin excitations RIXS can measure the dispersion of magnons. The energy of spin excitations tells us how strong the exchange interaction between Cu moments is, ie how stable the antiferromagnetic order. R. Coldea et al, Phys. Rev. Lett. 86, 5377 (2001). Until 2010 only neutrons could do that, with limitations due to tiny x-sections High resolution is necessary L. Braicovich, J. van den Brink, M. Moretti Sala, GG et al PRL (2010)

36 RIXS: paramagnons Interestingly RIXS has demonstrated that antiferromagnetism remains very strong even in doped, superconducting cuprates. This observation makes spin fluctuations the best candidate for Cooper pairing! L. Braicovich, J. van den Brink, M. Moretti Sala, GG et al PRL (2010) M. Le Tacon, GG, B. Keimer et al, Nat. Phys. 7, 725 (2011)

37 ERIXS at ESRF: full maps of magnons The detailed maps of spin excitations reveal why different families of cuprates have different max T c NdBa 2 Cu 3 O 6.1 CaCuO 2 Y.Y. Peng, GG et al, arxiv: (2016)

38 Paramagnons in Bi2201 UD15 Inelastic part Strongly damped collective spin excitations or e-h Stoner like excitations? OP33 (1/2,1/2) (0,0) (1/2,0) YY. Peng, GG et al, unpublished

39 R(I)XS: charge density waves RIXS was used to discover CDW in YBCO. Within a few months, CDW were observed, with RIXS and other techniques, in many other cuprates NBCO Tc=65K V pol, T=15K rlu rlu rlu rlu rlu rlu Energy loss (ev) Max intensity at Tc: CDW compete with SC RXS (and STM, XRD and NMR) has demonstrated that CDW are present in all underdoped cuprates. Are CDW related to Superconductivity? GG., M. Le Tacon, M. Minola, B. Keimer, L. Braicovich, et al Science 337, 821 (2012)

40 CDW and new generation RIXS Higher sensitivity reveals CDW in optimally doped Bi2201 Bi 2 (Sr,La) 2 CuO 6+δ

41 ERIXS: electron-phonon coupling The very high resolution of ERIXS at ID32 allows the detection of phonons with RIXS. This is a very direct way of measuring the e-ph coupling. The mapping of the e-ph coupling can clarify the role of CDW in superconductivity L. Braicovich, S. Johnston, J van den Brink, in preparation

42 More work on e-ph coupling Theoretical phonon intensity in RIXS Bond stretching A 1 (buckling) L.A. INS (LBCO) Phonons and CDW T. P. Devereaux, A. M. Shvaika, K. Wu, K. Wohlfeld, C.-J. Jia, Y. Wang, B. Moritz, L. Chaix, W.-S. Lee, Z.-X. Shen, G. Ghiringhelli, and L. Braicovich, PRX 6, (2016) L. Chaix, G. Ghiringhelli, Y. Y. Peng, M. Hashimoto, Y. He, S. Chen, K. Kummer, N.B. Brookes, B. Moritz, S. Ishida, Y. Yoshida, H. Eisaki, L. Braicovich, Z.-X. Shen, T. P. Devereaux, W.-S. Lee, submitted Exp. (40 mev BW) Theory Exp. (120 mev BW) Energy loss (ev) I(CDW)= NBCO UD Reznik, 0 arxiv Momentum transfer (q/π,0) Energy loss (ev)

43 Nature 518, 179 (2015) Conclusions

44 Thanks

45 A3 B1 Phonons in TiO 2 B2 B0 RIXS Intensity [arb.units] mev 2θ=150 ( q//=-0.17 π/a, qz=0.52 π/c) 2θ=120 ( q//=-0.10 π/a, qz=0.55 π/c) 2θ=90 ( q//=-0.03 π/a, qz=0.49 π/c) TiO 2 Insulating Resonant on B1- H polarization BW = 28 mev 300 mev 200 mev 100 mev 65 mev Elastic peak B1 excitation Different q Moser et al, PRL 115, (2015) Energy Loss [ev] M. Salluzzo, G.M. De Luca, YY Peng, R. Fumagalli, D. Betto, L. Braicovich, G. Ghiringhelli, unpublished

46 Soft x-ray spectrometer SAMPLE ILLUMINATED SURFACE INCIDENT BEAM 2D POSITION SENSITIVE DETECTOR AREA INCLUDED IN THE SPECTROMETER ACCEPTANCE EMITTED PHOTONS SPECTROMETER ENTRANCE SLIT SPECTROMETER GRATI NG

47 Longer instruments for better resolution CCD Sample Grating

48 AXES: 2.2 m ID12B and ID08 SAXES, SLS: 5 m ERIXS, ID32: 10 m

49 Resolving power: 40,000 at 1 kev 3 times better than previous record Collimating Mirror to increase horizontal acceptance ERIXS at ID32 Two CCD detectors Multi-layer mirror, to measure polarization of scattered photons Two Gratings, to optimize for resolution or efficiency