Transcript Document
Chern-09 Schegolev Memorial Conference 2009 Supercurrents in ferromagnets J. Aarts, M. S. Anwar Kamerlingh Onnes Laboratory, Leiden. I. The theoretical scene : - from LOFF state to odd-frequency triplets. II. The experimental scene : - Experiments with CrO2 . + discussions / experiments with T. Klapwijk (Delft), S. Goennenwein, F. Cheska (HMI- München) S/F hybrids – some history, proximity effect Chern-09 S/F multilayers, oscillation of Tc(dF) Theory : Radovic, .., Buzdin PRB 1991 Experiment : Nb / Gd multilayer Jiang, PRL 1995 Chern-09 I. F / S hybrids; inhomogeneous superconductivity in ‘weak F’ The 'LOFF' - state : pairing in presence of exchange field : not between +k↑ and -k↓. • Larkin & Ovchinnikov, Sov. Phys. JETP '65; • Fulde & Ferrell, Phys. Rev. '64 Characteristic : inhomogeneous pair density; e.g C-pair from S to F. In F (exchange field h), pair gains momentum 2h Q eiQx or 2h Q e iQx vF vF oscillates : cos(Qx) So, S induces in F : oscillatory damped pair density. Chern-09 Oscillations : length scale ξF F 1,2 and can change phase – by π D ( Eex2 ( k BT )2 )1/ 2 k BT Eex >> kBT , F DF Eex Eex 1 eV (Ni) ξF 1 nm Eex kBT ξF1 : decay ξF2 : period Need weak magnets for large ξF : Cu50Ni50 , Pd90Ni10 , etc. ξF → 10 nm S / F / S π - junction Chern-09 Consequences : oscillations in Tc(dF) and Ic(T) / Ic(dF) Nb / Co multilayer Nb / CuNi junction Obi, Phys C ‘99 Oboznov – Ryazanov, PRL ‘06 note small dF New development : odd-frequency triplets Chern-09 attractive interaction + exchange : spin singlet + spin triplet Pauli : singlet : spin odd, orbit even s , (d) ‘ = ‘ Nb, YBCO triplet : spin even, orbit odd p , (f) ‘ = ‘ Sr2RuO4 but, ‘Pauli’ = ‘equal times’ only. Using negative times / frequencies allows to circumvent this : triplet : spin even, ω odd, orbit even s : isotropic , (d) Matsubara : n n with f0(n) = f0(-n) and f1(n) = -f0(-n) The receipe for triplets : Spin mixing by different spin scattering at interface singlet |> - |> → m=0 triplet |> + |> Spin rotation by exchange field then also yields m=1 triplet |> and |> Chern-09 Volkov / Bergeret / Evetov • mix at interface, • rotate in domain wall • end with |> Possibly observed : Ho bridge in Al loop (Sosnin – Petrashov, PRL ’06) triplet singlet Triplet is not broken in F : Long range proximity ! Supercurrent Chern-09 Halfmetallic ferromagnet – mix and rotate at interface Eschrig, N Phys ‘ 08 • weak magnet does not introduce much m = 0 component. • rotation is by disordered interface moments. • Effect is long range, no spin flip in HFM. F DF k BT Special feature : π / 2 shift at each interface π – junction without thickness dependence Chern-09 Possible to measure ? Braude ’07 , Asano ‘07 ½ Φ0 by scanning SQUID LDOS by low-T STM S S F Δ0 zero-bias conductance peak Hilgenkamp / Kirtley 2006 Chern-09 Zero-bias conductance peak ? As in d-wave HTS Surface Andreev Bound State has zero-energy solution - ZBPC YBCO (110) sample Maarten v.Zalk Twente data Simon Kelly Leiden Leiden LT-UHV-STM, 300 mK, 8 T operational since fall ’08 (Federica Galli) Chern-09 Candidates for the triplet supercurrents : (La,Sr)MnO3 , CrO2 Found in CrO2 / NbTiN ? 100 nm Keizer, Klapwijk, Gupta et al, Nature 2006 F - films from Alabama ; S - contacts in Delft Note the biaxial anisotropy revisited in Leiden Chern-09 CrO2 – difficult as thin film • not by sputtering, PLD, MBE only CVD, and only on TiO2 modified with extra precursor heater • S contacts are not grown in-situ surface cleaning issue • intermediate thickness has biaxial anisotropy Chern-09 Morphology is subtle – depends on pretreatment TiO2 (HF etch) Basic properties reproduce well TiO2, Untreated TiO2, Treated [001] 1.0µm 1.0µm roughness of order 2 nm Basis for biaxial anisotropy. easy axis Bulk CrO2 : c Strained (TiO2) : b Relax : bi-axial (100 nm) c (200 nm Chern-09 2008 – new expts on Alabama and Leiden samples contacts by lift-off after Ar-etching the surface. CrO2 + (Nb,Ti)N and CrO2 + a-MoGe Problem is interface and / or magnetic stuff and / or something else. Try something different, grow on sapphire ….. CrO2 Cr2O3 Al2O3 no supercurrent Chern-09 CrO2 on Al2O3 [001] 60o 60o [001] a AFM 3c Chern-09 HREM Rabe – Güntherodt J.Phys.Cond.Mat ‘02 Note the Cr2O3 layer, and the columnar growth of the CrO2 Use different structure (larger) Chern-09 zero-bias resistance lift-off film disordered and rough I-V, current-biased Chern-09 200 J004 J005 dCrO2 = 100 nm IC (A) 150 104 A/cm2 100 compare Nb / CuNi 50 0 1 2 3 4 5 6 T (K) Ic (T) - 1 μm slit; two samples Dev. A : 1 mA 5 105 A/cm2 dCrO2 = 100 nm Chern-09 Ic (Ha) at 3 K, field in-plane, bridge compare CrO2 on TiO2 45 mT 80 mT • increase with Ha • Φ0 / 2 = 80 mT junction area : 1 μm dCrO2 junction area : 0.3 μm dCrO2 dCrO2 = 10 nm - smallish dCrO2 = 70 nm - nominally 100 nm Chern-09 Thouless-energy analysis S/N : Ic ~ T3/2 exp(-2πkBT / Eth)1/2 ; -8 Eth=72eV J004-1B J005-2A Eth=91eV eRNIc = 10.82 Eth ETh 80 μeV InIC-3/2InT -10 = 2 μΩcm RN = 0.2 Ω -12 (d = 10 nm) or RN = 20 mΩ (d = 100 nm) -14 1.0 1.5 2.0 T 1/2 1/2 (K ) 2.5 Icmax = 4 mA (d = 10 nm) or Icmax = 40 mA (d = 100 nm) Ic (too) small - grain boundaries ? Chern-09 Conclusion and outlook : • looks like confirmation of the 1st report • promise of (very) long range effects (1 μm at 4 K) • promise of a novel π – junction (no thickness dependence) In progress • smaller gaps (increase Ic) • contact CrO2 strips instead of film → rings • Ic to 300 mK → Eschrig 2008 Main questions still not answered …. • what constitutes a magnetically active interface • what is the role (if any) of the magnetic anisotropy ? Chern-09 Alabama – Delft sample