Transcript Slide 1
Beyond ferromagnetic spintronics: antiferromagnetic I-Mn-V semiconductors Tomas Jungwirth Institute of Physics in Prague & University of Nottingham Spintronics ← relativistic quantum physics Kvantová relativistická fyzika p2 E 2m i (r , t ) t 2 2 (r , t ) 2m r 2 Spintronics ← relativistic quantum physics Kvantová relativistická fyzika E m c , m m0 2 1 / (1 v / c ) 2 2 Spintronics ← relativistic quantum physics Kvantová relativistická fyzika Spin-orbit coupling } } Ultra-relativistic particles with spin (neutrino): Weaker but also present in electrons in solids E cp s Electron has spin & charge → magnetic moment | m | B Collective behavior of spins due to Coulomb interaction → magnetism Provides sensitivity to weak external fields & yields strong electrical signals Electron has spin & charge → magnetic moment | m | B Collective behavior of spins due to Coulomb interaction → magnetism Provides sensitivity to weak external fields & yields strong electrical signals Electron has spin & charge → magnetic moment | m | B Collective behavior of spins due to Coulomb interaction → magnetism Provides sensitivity to weak external fields & yields strong electrical signals ... and memory Spintronic magnetoresistance effects in metals Bulk AMR TMR (GMR) Lord Kelvin 1857 Fert, Grünberg et al. 1988 Magnetic RAM HDD read-head sensors 2 R(m) R(m) R(m ) First spintronic devices Poor scalability to small dimensions & small MR (subtle spin-orbit origin) R(m) R(m) R(m) Current spintrnic devices Interface effect → nanoscale in nature & large MR (robust ferromagnetic origin) Towards semiconductor spintronics FM semiconductors Ohno et al. Science’98, Dietl et al PRB’00, Jungwirth, MacDonald et al PRB’99 Archetypical material (Ga,Mn)As: favorable FM and spin-orbit coupled bands & semiconductor nano-fabrication → revived interest in spin-orbit phenomena like AMR in nanostructures Huge (~1000%) AMR-type effects in (Ga,Mn)As nanostructures → (m) Electrical control of spintronics Positive & negative MR VG1 VG2 → B (T) → rotating m Spintronic control of electronics p-type & n-type transistor → m1 Wunderlich, Irvine, Jungwirth et al. PRL’06, Schlapps, Weiss et al. PRB’09 → m2 Limitations of ferromagnetic semiconductor (Ga,Mn)As Well behaved Itinerant ferromagnet but... (Ga,Mn)As ...FM at huge dopings > 1% (> 1020 cm-3 ) → more of a low-density metallic alloy Tc below room-T ( 190K) (Ga,Mn)As Tc Novák, Jungwirth et al. PRL ’08 AMR-type effects predicted and observed in high-Tc FM metal nanostructures Theory predictions Shick, Jungwirth et al. ‘06 Wunderlich, Jungwirth, Shick et al. ’06 Confirmed by experiments Gao, Tsumbal, Parkin et al. ’07 Park,Wunderlich, Jungwirth et al. 08 Bernand-Mantel, Fert et al. ‘09 Pt AlOx Pt/Co cobalt Maximizing the anisotropy phenomena in metals → spintronics in the AFMs AFM metal MnIr spontaneous moment spin-orbit coupling FM AFM 2 R( m ) Magnetic and magneto-transport anisotropy effects present in AFMs with spin-orbit equally well as in FMs Shick , Wunderlich, Jungwirth, et al., PRB‘10 Can AFMs resolve the problem of high-T SEMICONDUCTOR spintronics? Jungwirth, Novak, et al., preprint ‘10 Eexchange competing with Egap in FM-SCs No Eexchange competing with Egap in AFM-SCs Eexchange Egap Strong FM exchange spitting turns the system into metal EFermi Much easier to realize strong AFM-SC than FM-SC Magnetic (FM & AFM) SCs derived from common 8-valence non-magnetic SCs II III Mn (d5 s2) Fe Eu (f7 s2) Gd II III IV V (pnictides) VI (chalcogenides) Zn, Cd, .. Al, Ga, .. Si, Ge, .. N, P, As, .. O, S, Se, Te, .. Si Si 2 group-IV Si per elementary cell → 8 (sp) valence electrons Magnetic (FM & AFM) SCs derived from common 8-valence non-magnetic SCs II III Mn (d5 s2) Fe Eu (f7 s2) Gd II III IV V (pnictides) VI (chalcogenides) Zn, Cd, .. Al, Ga, .. Si, Ge, .. N, P, As, .. IV: no magnetic SC analogue O, S, Se, Te, .. Magnetic (FM & AFM) SCs derived from common 8-valence non-magnetic SCs II III Mn (d5 s2) Fe Eu (f7 s2) Gd II III IV V (pnictides) VI (chalcogenides) Zn, Cd, .. Al, Ga, .. Si, Ge, .. N, P, As, .. O, S, Se, Te, .. IV: no magnetic SC analogue 1 proton transfer IV Si Si III-V Magnetic SCs derived from common 8-valence non-magnetic SCs II III Mn (d5 s2) Fe Eu (f7 s2) Gd II III IV V (pnictides) VI (chalcogenides) Zn, Cd, .. Al, Ga, .. Si, Ge, .. N, P, As, .. O, S, Se, Te, .. IV: no magnetic SC analogue III-V: FeAs – SC, AFM TN=77K GdN – SC, FM Tc=72K (Ga,Mn)As – low-density metal, FM Tc<190K Lower moment Fe (Gd) less favorable than high moment Mn → II-VI intrinsic magnetic SCs Magnetic SCs derived from common 8-valence non-magnetic SCs II III Mn (d5 s2) Fe Eu (f7 s2) Gd II III IV V (pnictides) VI (chalcogenides) Zn, Cd, .. Al, Ga, .. Si, Ge, .. N, P, As, .. O, S, Se, Te, .. IV: no magnetic SC analogue III-V: FeAs – SC, AFM TN=77K GdN – SC, FM Tc=72K (Ga,Mn)As – low-density metal, FM Tc<190K Magnetic (FM & AFM) SCs derived from common 8-valence non-magnetic SCs II III Mn (d5 s2) Fe Eu (f7 s2) Gd II III IV V (pnictides) VI (chalcogenides) Zn, Cd, .. Al, Ga, .. Si, Ge, .. N, P, As, .. O, S, Se, Te, .. IV: no magnetic SC analogue III-V: FeAs – SC, AFM TN=77K GdN – SC, FM Tc=72K (Ga,Mn)As – low-density metal, FM Tc<190K II-VI: MnO, MnS, MnSe, MnTe - SC, AFM TN ~ 100 - 300K EuO, EuS – SC, FM Tc<70K EuSe, EuTe - SC, AFM TN<10K Larger more ionic bonds weaken magnetic interactions in II-V‘s All III-V and II-VI magnetic SCs have low transition-T Can we make high moment (Mn) and smaller lattice (pnictides) intrinsic SC? Magnetic (FM & AFM) SCs derived from common 8-valence non-magnetic SCs II III Mn (d5 s2) Fe Eu (f7 s2) Gd I (AM) Li, Na,.. (TM) Cu, Ag, .. II III IV V (pnictides) VI (chalcogenides) Zn, Cd, .. Al, Ga, .. Si, Ge, .. N, P, As, .. O, S, Se, Te, .. Magnetic (FM & AFM) SCs derived from common 8-valence non-magnetic SCs II III Mn (d5 s2) Fe Eu (f7 s2) Gd II III IV V (pnictides) VI (chalcogenides) Zn, Cd, .. Al, Ga, .. Si, Ge, .. N, P, As, .. O, S, Se, Te, .. I (AM) Li, Na,.. (TM) Cu, Ag, .. I-II-V: LiMnAs, NaMnAs, LiMnP, LiMnSb... - AFM TN >> room T Bronger et al., Z. among. allg. Chem. ’86 Magnetic (FM & AFM) SCs derived from common 8-valence non-magnetic SCs II III Mn (d5 s2) Fe Eu (f7 s2) Gd II III IV V (pnictides) VI (chalcogenides) Zn, Cd, .. Al, Ga, .. Si, Ge, .. N, P, As, .. O, S, Se, Te, .. I (AM) Li, Na,.. (TM) Cu, Ag, .. I-II-V: LiMnAs, NaMnAs, LiMnP, LiMnSb... - AFM TN >> room T Bronger et al., Z. among. allg. Chem. ’86 III-V I-II-V Twin SCs I-Mn-V Magnetic (FM & AFM) SCs derived from common 8-valence non-magnetic SCs II III Mn (d5 s2) Fe Eu (f7 s2) Gd II III IV V (pnictides) VI (chalcogenides) Zn, Cd, .. Al, Ga, .. Si, Ge, .. N, P, As, .. O, S, Se, Te, .. I (AM) Li, Na,.. (TM) Cu, Ag, .. I-II-V: LiMnAs, NaMnAs, LiMnP, LiMnSb... - AFM TN >> room T Bronger et al., Z. among. allg. Chem. ’86 No report on electronic structure of AFM I-Mn-V: Are they SCs? No report on MBE growth of group-I compounds: Can they be grown as single-crystal epilayers? I-Mn-V MBE growth of I-Mn-V: LiMnAs on nearly lattice matched InAs Li MnAs 4.27A 4.28A InAs In situ RHEED [110] In situ optical reflectivity LiMnAs [-110] InAs cap MnAs growth drection log(intensity) Sharp 2D cubic single-crystal growth 3.5 ~2 LiMnAs 1000 1400 1200 wavelength (nm) Fabry-Perot oscillations → semiconductor ... poor growth of control umatched MnAs profile (nm) Ex situ profile 200 LiMnAs 100 substrate 0 200 x 400 (m) 600 800 X-ray diffraction log(intensity) Li MnAs 4.27A All LiMnAs crystal peaks observed 4.28A InAs Fully tensile strained on InAs (0.2% increase of LiMnAs volume) Li MnAs X-ray diffraction LiMnAs [110] InAs [100] InAs Expected 45o rotation of LiMnAs with respect to the InAs substrate Ex situ optical transmission LiMnAs Li:InAs MnAs Mrem (104 emu) IT/I0 InAs Squid magnetization MnAs LiMnAs temperature (K) energy (meV) Transparent at least up to InAs band-gap Magnetization consistent with compensated AFM moments in LiMnAs upto studied 400K MnAs M (104 emu) Consistent with in situ Febry-Perot oscillations and compare with non-transparent metal MnAs Mn S=5/2 LiMnAs H (T) Compare with FM MnAs with same amount of Mn Ab initio theory Stoichiometric I-Mn-V are strong AFMs & intrinsic semiconductors LDA Magnetic and correlated Mn d-states mixed near band gap → low √ (refractive index), strong and gatable magnetic anisotropy effects AFM semiconductors for spintronics 1. Electrically gatable magnetic and magneto-transport anisotropy effects FM AFM Feasible to rotate magnetic easy-axis electrically in high-doped (Ga,Mn)As → should be much more accessible in intrinsic SCs I-Mn-V AFM semiconductors for spintronics 2. Exchange-biasing AFM with embeded conventional semiconductor devices Fixed by exchangebiasing AFM Transistor directly in the AFM layer Discrete spintronic and transistor elements in current MRAM Opto-electronics directly in the AFM layer Conclusions FM SCs (GaMnAs) favorable model spintronic systems but low transition T AFM I-Mn-V compounds: - Simplest magnetic counterparts to conventional SCs with high transition T - We showed that they are semiconductors and that the group-I alkali metal compounds can be grown by MBE as high quality single-crystal epilayers - Admixture of magnetic d-states yields unconventional SC properties and theory predicts very strong and gatable spintronic responses Prospect for high-T semiconductor spintronics but first sytematic materials research needs to be completed Institute of Physics ASCR, Prague Vít Novák, Miroslav Cukr , Jan Mašek, Alexander Shick, František Máca,Petr Kužel, et al. University of Nottingham Tom Foxon, Richard Campion, Bryan Gallagher, et al. Charles University, Prague Xavi Marti, Petra Horodyská, Václav Holý, Petr Němec, et al. Hitachi & Cavendish Laboratories at Cambridge Jorg Wunderlich, Andrew Irvine et al. Texas A&M and University of Texas Jairo Sinova, Allan MacDonald, et al.