Optical Control in Semiconductors for Spintronics and ... › ece › Affil › archive › 2003 ›...

Optical Control in SemiconductorsOptical Control in Semiconductorsfor Spintronicsfor Spintronics

andandQuantum Information ProcessingQuantum Information Processing

Jun Kono

Department of Electrical & Computer Engineering,Rice University

October 8, 2003

Supported by NSF DMR-0325474 (ITR), NSF DMR-0134058 (CAREER), NSF INT-0221704, and DARPAMDA972-00-1-0034 (SpinS)

ECE Affiliates Meeting

October 8, 2003 J. Kono

Our Spin/Q.I.P. TeamOur Spin/Q.I.P. Team

Jigang Wang Ajit Srivastava Xiangfeng Wang Rahul Srivastava Giti Khodaparast

Dave ReitzeChris Stanton Young-Dahl Jho

Lu Sham Hiro Munekata

Univ. of Florida

UCSD Tokyo Inst.of Technology


Need for Need for QuantumQuantum Technologies Technologies

ß Miniaturization – approaching a physical limit

ß Quantum effects – statistical, fuzzy, strange – unavoidable

ß Novel quantum technologies being sought ‡ better performanceand new functionality and multi-functionality

ß Spin-based electronics, quantum electronics based on quantumcoherence, interference, and entanglement, … etc.

“Quantum physics holds the keyto the further advance ofcomputing in the postsilicon era.”

- J. Birnbaum and R. S. Williams

“Coherent spin packets may offergenuine quantum devicesthrough their wave-like properties.”

- D. D. Awschalom


OutlineOutline

n Semiconductor ‘Spintronics’n Towards Solid-State Realization of Quantum

Information ProcessingOur Approach: Ultrafast Optical ControlOur Recent Discoveries

Ultrafast Photoinduced Softening (UPS) in aFerromagnetic SemiconductorUltrafast Photoinduced Transparency (UPT) using theDynamic Franz-Keldysh Effect (DFKE)

Summary


Emerging Technologies forEmerging Technologies forSolid-State Information ProcessingSolid-State Information Processing

Spintronics

ß Use spinsspins in solid statedevices

ß Improve informationprocessing

ß Add new functionalities

Quantum InformationProcessing

ß Devise and implementquantum-coherent

strategies for computationcomputationand communicationcommunication

Use discrete, quantized degrees of freedom in a physicaldevice to perform information processing functions.


Magneto-ElectronicsMagneto-Electronics

Magnetic tunneling junction(MTJ) or “spin valve” ‡Nonvolatile MRAM:“Microchips that neverforget ”

S. Parkin (1990)Compatibility with Si andGaAs ‡ next phase:semiconductor spintronics

GMR ‡ read-out heads in hard drives

1st generation spintronic devices based onferromagnetic metals – already in commercial use


Recent Discoveries in SemiconductorsRecent Discoveries in Semiconductors

n A room temperature, optically induced, verylong lived quantum coherent spin state insemiconductors that responds at Terahertzwith no dissipation and can be transported bysmall electric fields (UCSB).

n Ferromagnetism in semiconducting GaMnAsat 120K (Japan, Europe, U.S.A.).

DARPA ‘Spins in Semiconductors’Program (2000 – present)


Spin-Enhanced and Spin-Spin-Enhanced and Spin-Enabled ElectronicsEnabled Electronics

• Quantum Spin Electronics– Tunneling/transport of quantum confined spin states

– Spin dependent resonant tunneling devices and spin filtering

– Spin FETs (“spin gating”)

– Spin LEDs, electroluminescent devices, and spin lasers

• Coherent Spin Electronics– Optically generated coherent spin states and coherent control of

propagating spin information - optical encoders and decoders

• Quantum Information Processing– Qubits using coherent spin states a|0> + b|1>, a2 + b2 = 1

– Spin based quantum computing, teleportation, code breaking andcryptography


Quantum Information ProcessingQuantum Information Processing

- Coherent superposition: 10 ba +=y

‡ Inherent parallelism ‡ can solve problems that arecomputationally too intensive for classical computers

- 2 examples of ‘quantum algorithms’:Shor’s factorization (1994)Grover’s search (1997)

- Quantum entanglement, EPR pair, quantum teleportation

- Physical Implementations:Trapped ions (1995)Cavity QED (1995)Bulk NMR (1997), …

Proof-of-principledemonstrations, butnot scalable


Toward Toward Solid-StateSolid-State Realization of QIP Realization of QIPThe Race is on!The Race is on!

Semiconductors vs. Superconductors

Spin Qubits vs. Charge Qubits

Electrical Manipulation vs. Optical Manipulation

• Intensive search for realistic approaches to building aquantum computer

• Solid-state systems offer a much greater degree ofcontrol over design and fabrication, necessary forconstructing large-scale devices

Cooper Pairs vs. Flux QubitsElectron Spin vs. Nuclear Spins


Decoherence Decoherence ProblemProblem

T : decoherence timet : operation timeR = T/t : figure of merit

How can we increase T and/or decrease t?

• Coherent states are very easily damaged by uncontrolledinteractions with the environment – decoherence

• Unavoidable decoherence will cause the quantuminformation to decay ‡ main obstacle

• Decoherence causes a collapse of the superposition stateinto a single eigenstate ‡ loss of parallelism


To develop novel ultrafast optical methodsin semiconductors that may find application inspintronics or quantum information sciencethrough coherent light-matter interactionsinvolving ferromagnetism, band structures,lattice vibrations, and excitons

Our GoalOur Goal

e-


Ultrafast Ultrafast Photoinduced SofteningPhotoinduced Softening(UPS) in InMnAs(UPS) in InMnAs

J. Wang, G. A. Khodaparast & J. Kono

Rice University

T. Slupinski, A. Oiwa & H. Munekata

Tokyo Institute of Technology


• Ultrafast demagnetization (~ hundred fs)and slow recovery

• Possible application to ultrafast magneto-optical recording

E. Beaurepaire et al., PRL 76, 4250 (1996); PRB 58, 12134 (1998).

Ni CoPt3

Ultrafast Optics in Ferromagnetic Ultrafast Optics in Ferromagnetic MetalsMetals


Low-temperature MBEgrown III1-xMnxV:

• InMnAs: Tc < 60 K

• GaMnAs: Tc < 170 K

Mn ions (Mn2+) = acceptors& local magnetic moments(3d5, S = 5/2)

Mn-Mn exchange:hole mediated

Carrier density tuning

External control of ferromagnetism

III-V Ferromagnetic SemiconductorsIII-V Ferromagnetic Semiconductors

October 8, 2003 J. KonoAdvantages ofAdvantages of

Ferromagnetic SemiconductorsFerromagnetic Semiconductorsover over Ferromagnetic MetalsFerromagnetic Metals

• Ultrafast pump ‡ primarily increasescarrier density (rather than carriertemperature)

• Created carriers interact with Mn ions ‡enhance Mn-Mn exchange interaction

• Circular-polarized pump ‡ spin polarizedcarriers

• Low-T MBE growth ‡ ultrashort lifetimes


Ultrafast Carrier Ultrafast Carrier DDynamicsynamics in InMnAs in InMnAs

(1) carriertrapping

(~ 2 ps)

(2) Carrierrecombinationof trappedcarriers

-3

-2

-1

0

1

2

?R

/R (

%)

200150100500

Time Delay (ps)

Pump = 1260 nmProbe = 775 nm

(1)

(2)


-6x10-5

-4

-2

0

2

4

6

MC

D S

igna

l (V

)

-0.4 -0.2 0.0 0.2 0.4

Magnetic field (T)

55 K 45 K 35 K 20 K 10 K

In0.91Mn0.09As(25nm)/GaSb(820nm) on GaAs(100)Tc = 55 K

Magneto-optical Kerr Effect (MOKE)Magneto-optical Kerr Effect (MOKE)


l Selective pumpingof InMnAs by fs MIRpulses

l Photogenerated,transient spin-polarized carriers

l Probe time-dependentferromagnetism

Two-Color Pump & ProbeTwo-Color Pump & Probe

InMnAs GaSb

MIRMIRpumppump

~150 fs~150 fs

NIR NIR probeprobe

holes

V

C

m


-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

MO

KE

Sig

nal (

deg.

)

6040200Delay Time (ps)

_K

_K

B (T)

T = 16 KTc = 55 KH = 0 OePump: 2 mmProbe: 775 nm

0.4

0.2

0.0

-0.2

-0.4-0.2 -0.1 0.0 0.1 0.2

0.4

0.2

0.0

-0.2

-0.4

-0.2 -0.1 0.0 0.1 0.2

J. Wang et al., J. Supercond. 16, 373 (2003); Physica E, in press; cond-mat/0305017

Ultrafast Photoinduced MOKEUltrafast Photoinduced MOKE


t ~ -4 ps t ~ 0 ps t ~ 11 ps

n Loop shrinks horizontally and then comes back!

n First demonstration of ultrafast optical manipulationof coercivity

-4

-2

0

2

4

MOK

E Si

gnal

(10-6

V)

-0.2 -0.1 0.0 0.1 0.2Magnetic Field (T)

-78

-76

-74

-72

-70

-68

-66

MOK

E Si

gnal

(10-6

V)

-0.2 -0.1 0.0 0.1 0.2Magnetic Field (T)

-4

-2

0

2

4

MOK

E Si

gnal

(10-6

V)

-0.2 -0.1 0.0 0.1 0.2Magnetic Field (T)

J. Wang et al., J. Supercond. 16, 373 (2003); Physica E, in press; cond-mat/0305017

Ultrafast Photoinduced SofteningUltrafast Photoinduced Softening


NonthermalNonthermal MagnetoopticalMagnetooptical Recording Recording

Hc0

B

M

Hcmod < Hex < Hc

0

B

M~100 fs

Hcmod

t = 0

t > 0

Spin flipping ‡ ultrafast information recording

Dynamic Franz-Keldysh EffectDynamic Franz-Keldysh Effect(DFKE) in GaAs:(DFKE) in GaAs:

Optical absorption in AC-driven solidsOptical absorption in AC-driven solids

A. Srivastava, R. Srivastava, J. Wang, and J. Kono

Department of Electrical & Computer Engineering,Rice University, U.S.A.

1. Predictions

2. Observations

3. Interpretation

4. Significance

MIRpump

c

v

NIRprobe

Ultrafast bandgap engineering


Wiggling Wiggling EElectronlectron in a in a LLaser aser FFieldield

Classical energy Classical energy UUpp

e-

potential iveponderomot : 42

11K.E.

2

20

2

0

2p

TU

m

Eedtmv

T≡=˜

¯

ˆÁË

Ê= Ú w

( )

( ) tmeE

tv

tEtE

ww

w

sin

cos

0

0

-=fi

=

– wiggly motion

Quantum energyQuantum energy wh

transitionpU<<wh

pU>>wh

: DC-FKE

: Multiphoton


Time-Periodic Potentialvs.

Space-Periodic Potential

a0

Relevant length scale:

Bloch Electron Bloch Electron in a in a LLaser aser FFieldield

a

whªpU

20

wm

eEa =

ww ªB 0aa ª

fl Oscillation amplitude

)/( 00 hEeaB =w


How to realize How to realize aa ~ ~ aa00

022//aaEIgwl=µµ

0.1 1 10 100

1E7

1E8

1E9

1E10

1E11

1E12

1E13

1E14

1E15

m* = 0.07m0

Req

uire

d In

tens

ity (

W/c

m2

)

Wavelength (mm)

1=g

Ti:S

apph

ireN

d:Y

AG

OPA

FEL

22334pUeEImlww=µhh

easier at longerwavelengths

Intense MIRfs pulses


-1.0 -0.5 0.0 0.5 1.00.00.20.40.60.81.01.21.41.6

with pump without pump

Tra

nsi

tio

n r

ate

(a.u

)

hW - Eg (in units of hw)

Blue shiftBlue shift of band edge ‡transparency

Exponentialtail

Dynamic Franz-Keldysh Dynamic Franz-Keldysh EEffectffectY. Yacoby, PR 169, 610 (1968).

0aa ª


Photoinduced Photoinduced TTransparencyransparency~40%~40% photoinduced photoinduced transparency transparency

Band edge of GaAsAbsorption below band edge

GaAs film

Room temp.

Pump energy = 138 meV


DFKE Simulation for a GaAs FilmDFKE Simulation for a GaAs Film1.0

0.8

0.6

0.4

0.2

0.0

Tra

nsm

ittan

ce

no pump pumped

-0.6

-0.4

-0.2

0.0

0.2

T ?

T0

1.551.501.451.401.351.30

Energy (eV)

• Undoped GaAsfilm (2.7 mm thick)

• Multiple-reflectionincluded

• lpump = 9 mm (138meV)

• Ipump = 1010 W/cm2

ll InducedInducedtransparencytransparency


DFKE: DFKE: ConclusionsConclusions

n Intense MIR laser fields can coherentlymodify electronic states in solids throughnon-resonant pumping

n No sample damage, no real carriers

n Ultrafast transmission quenching belowband edge observed

n First observation of induced transparencyabove the band edge

n Main features of observations qualitativelyagree with theory


Accomplished:

• Photogenerated transient carriers ‡ modify magneticproperties in InMnAs ‡ first demonstration of ultrafastsoftening: Hc (coercivity) decreases (‘hard’ ‡ ‘soft’)

• Intense and coherent midinfrared radiation modified bandstructure through DFKE ‡ first observation of ultrafastphotoinduced transparency

SummarySummaryUltrafast Optical Control in SemiconductorsUltrafast Optical Control in Semiconductors

In Progress:

• Demonstration of ultrafast photoinduced magnetizationreversal

• Transient modifications of Tc and photoinduced transientpara- to ferromagnetic transition


Realization of Spin-Based DevicesRealization of Spin-Based Devices

Technical issuesTechnical issues

• How strongly can one create carriers of agiven spin?

• How long can one sustain the spinpolarization?

• How can one modulate or control the spin?

• How sensitively can one detect the spin?


Decoherence Decoherence ProblemProblem

T : coherence timet : operation timeR = T/t : figure of merit

Factoring a 4-bit number using Shor’s algorithmRequires ~2 x 104 gate operations on 20 qubits ‡R has to be > 4 x 105

• Coherent states are very easily damaged by uncontrolledinteractions with the environment – decoherence

• Unavoidable decoherence will cause the quantuminformation to decay, thus inducing errors in thecomputation

• Decoherence occurs rapidly in complex big systems, whichis why we never observe macroscopic superpositions


Ultrafast Optics in Ferromagnetic Ultrafast Optics in Ferromagnetic MetalsMetals

Ni and Co: E. Beaurepaire et al., Phys. Rev. Lett. 76, 4250 (1996). M. Aeschlimann et al., Phys. Rev. Lett. 79, 5158 (1997). A. Scholl et al., Phys. Rev. Lett. 79, 5146 (1997). J. Hohlfeld et al., Phys. Rev. Lett. 78, 4861 (1997). J. Güdde et al., Phys. Rev. B 59, R6608 (1999). B. Koopmans et al., Phys. Rev. Lett. 85, 844 (2000).

CoPt3: G. Ju et al., Phys. Rev. B 57, R700 (1998). E. Beaurepaire et al., Phys. Rev. B 58, 12134 (1998). L. Guidoni et al., Phys. Rev. Lett. 89, 017401 (2002).

Ultrafast Ultrafast ppump ump ‡‡ Electronic heating Electronic heating‡‡ Microscopic understanding elusive Microscopic understanding elusive


S. Koshihara et al., PRL 78, 4617 (1997); A. Oiwa et al., APL 78, 518 (2001).

CW Optical Control of FerromagnetismCW Optical Control of Ferromagnetism

n Light-induced ferromagnetism

n Light-induced coercivity decrease

n Persistent photoeffect


InMnAs GaSb

~0.8 eV

~0.4 eV

• Electron-holeseparation

• Hole densityincreasespersistently

• Slow

• Not reversible

Persistent Persistent PhotoeffectPhotoeffect

m


= tunable source of intense MIR pulses

MIR3 mm to 18 mmup to ~ 3 mJ

140 fs, 1 mJ

775 nm Ti:SRegenerative

Amplifier

up to 100 mJ

signal + idler signal1.1 mm to 1.6 mm

idler1.6 mm to 3 mm

RegenRegen + OPA + DFG + OPA + DFG

Optical ParametricAmplifier

DifferenceFrequencyGenerator

PRB 65, 121307(R) (2002)PRL 86, 3292 (2001) PRL 85, 3293 (2000)


Time Delay

MO

KE

Sig

nal q

t1t2t3

MO

KE

Sig

nal

Time Delay

DR

/R

H

Experimental Experimental SSetupetup


VE

CE

FE

InMnAsx < 9.5%

GaSb

~0.8 eV

~0.4 eV

holes

9-25 nm l Type-II broken gapheterostructures withAlGaSb

l First ferromagnetic p-type films, Tc = 7 K(1991), Tc = 35 K in p-InMnAs/GaSb (1993)

l Light-inducedferromagnetism (1997)

l Electrical tuning offerromagnetism (2000)

InMnAs-Based HeterostructuresInMnAs-Based Heterostructures


Carrier Carrier DDynamicsynamics in InMnAs in InMnAs

(1) carrier trapping

(~ 2 ps)

(2) recombinationof trappedcarriers

Pump 2Pump 2 mmm, 0.5m, 0.5 mW mWProbe 775 nmProbe 775 nm

V.B

C.B

(1)(2)

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

DR

/R (

%)

20151050

Time Delay (ps)

(1)

(2)


-8x10-5

-6

-4

-2

0

2

4

MO

KE

Sig

na

l (V

)

840-4

Time delay (ps)

-1

0

1

x10-5

-0.2 -0.1 0.0 0.1 0.2

Magnetic Field (T)

8

7

x10-5

-8

-7

x10-5

s+ pumping

s- pumping

20K

s+ pumping

s- pumping

t = 0

t = 0

t = -4 ps

Pump-Polarization DependencePump-Polarization Dependence


-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

Pho

to-in

duce

d M

OK

E S

igna

l (V

)

6420-2Time Delay (ps)

s+ pumping

s- pumping

122 K

MOKE Dynamics above MOKE Dynamics above TTcc

‡ Not related to ferromagnetism‡ This is due to carrier spin


What determines coercivity? -- Anisotropy & Exchange

CWCW

-8

-7

MO

KE

Sig

nal

0.20.10.0-0.1-0.2

Magnetic Field (T)

-1

0

1

(10 -

5 V

)

UltrafastUltrafast

• Anisotropy (K) does NOTchange with carrier density

• Exchange increases

Decreased domain wall energy‡ Smaller coercivity

22sincosiiijijEKJSqf=-Â

Origin of Ultrafast SofteningOrigin of Ultrafast Softening


Ultrafast Photoinduced SofteningUltrafast Photoinduced Softening


n Curie temperature (Tc) of InMnAs below roomtemperature ‡ explore new materials (e.g.,GaMnN)

n Small band gap of InMnAs requires powerfulmid-infrared (MIR) laser pulses ‡ requiresdevelopment of compact MIR source or largeband gap materials (e.g., GaMnN)

n Kerr rotation angle of InMnAs small ‡ newmaterials or sophisticated opticalarrangements necessary

LimitationsLimitations


E-field-induced changes in optical absorption near the E-field-induced changes in optical absorption near the bandedgebandedge

rEeÇrr

⋅-='Additional term in the Hamiltonian

Exponential for positive arguments

Oscillations fornegativearguments

Exponential tail below bandgap

Oscillationsabovebandgap

Franz-Keldysh Franz-Keldysh EEffectffect

absorption

wavefunction

dc fielddc field


Unusual Power DependenceUnusual Power Dependence

Dec

reas

ing

inte

nsi

ty

Incr

easi

ng

eff

ect!

!

Below band gap absorption in bulk GaAs at 300 KBelow band gap absorption in bulk GaAs at 300 K

0.14

0.12

0.10

0.08

0.06

0.04

0.02

0.00

Tra

nsm

issi

on

1.3901.3801.3701.360

Photon Energy (eV)

4.7 mm 6 mm 8 mm 9 mm


Extent of Induced AbsorptionExtent of Induced AbsorptionPolycrystalline ZnSe, 3.5 mm driving field, g ~ 1

0.5

0.4

0.3

0.2

0.1

0.0

tran

smis

sion

2.62.42.22.01.8photon energy (eV)

Induced absorption extends almost 1 eV below Eg!


DFKE is adramatic effect

Other Strong-Field-Induced EffectsOther Strong-Field-Induced Effectsin Semiconductorsin Semiconductors

DC Franz-Keldysh effect AC Stark effect

shift ~ only a few meV at 100 kV/cm !

B.G. Yacobi et al.A. Mysyrowicz et al.


Based on: Y. Yacoby, Phys. Rev. 169, 610 (1967).

DFKE Simulation for GaAsDFKE Simulation for GaAs

2.5

2.0

1.5

1.0

0.5

0.0

?

(106

cm-1

)

1.601.551.501.451.401.351.30

Energy (eV)

no pump pumped

GaAs? pump = 9 ?m (h? = 138 meV)

Ipump = 1010 W/cm2 (Up~ h?)

Blue shift of band edge ‡Inducedtransparencytransparency

Inducedabsorption


Pump Pump WWavelength avelength DDependenceependence

Band edge of GaAs

GaAs film

Room temp.


ExperimentsExperiments

• A. H. Chin, J. M. Bakker, and J. Kono, Phys. Rev. Lett. 85, 3293 (2000).• A. H. Chin, O. G. Calderón, and J. Kono, Phys. Rev. Lett. 86, 3292 (2001).• O. G. Calderón, A. H. Chin, and J. Kono, Phys. Rev. A 63, 053807 (2001).• M. A. Zudov, J. Kono, A. P. Mitchell, and A. H. Chin, Phys. Rev. B 64,

121204(R), (2001).• A. H. Chin, J. Kono, and G. S. Solomon, Phys. Rev. B 65, 121307(R),

(2002).• A. Srivastava and J. Kono, in: Quantum Electronics and Laser Science

Conference, OSA Technical Digest (Optical Society of America,Washington DC, 2003), QFD2.

Optical Control in Semiconductors for Spintronics and ... › ece › Affil › archive › 2003 ›...

Documents

Transcript of Optical Control in Semiconductors for Spintronics and ... › ece › Affil › archive › 2003 ›...