Quaternionic states of matter from synthetic gauge...

Post on 21-Mar-2020

1 views 0 download

Transcript of Quaternionic states of matter from synthetic gauge...

Congjun Wu University of California, San Diego

Quaternionic states of matter from synthetic gauge fields

ISCAP Taiyuan,

Jun 16, 2014

Collaborators: Yi Li (UCSD Princeton), X. F. Zhou (USTC, Hefei)

2e

1e

S

Q-BEC with Hopf invariant 3D Q-analytic Landau level

2

AcknowledgementCollaborations on related topics: Vincent Liu, Zi Cai, Luming Duan, Yue Yu, K. Intrilligator, S. C. Zhang.

Helpful discussions with Jorge Hirsch, Kazuki Hasebe, T. L. Ho, Jiang-ping Hu, Nai Phuan Ong, Cenke Xu, Kun Yang, Yue Yu, S. C. Zhang, Fei Zhou

Support from NSF-DMR, AFOSR.

Ref. 1) Y. Li, X. F. Zhou, C. Wu, arxiv1205.2162.2) Y. Li, C. Wu, PRL 110, 216802 (2013), arxiv:1103.5422.3) Y. Li, Xiangfa Zhou, C. Wu, Phys. Rev. B 85, 125122 (2012).4) Review on synthetic SO coupling: X. F. Zhou, Y. Li, Z. Cai, C. Wu, J

Phy. B 46, 134001 (2013).

Other related work1) C. Wu, I. Mondragon-Shem, X. F. Zhou, Chin. Phys. Lett. 28, 097102, 2011(arXiv:0809.3532).

2) Y. Li, S. C. Zhang, C. Wu, PRL 111, 186803(2013).3) Y. Li, K. Intrilligator, Y. Yu, C. Wu, PRB 85, 85132(2012).

Outline

• Real numbers complex numbers quaternions (Q).

• Bosons: real (positive) BEC complex BEC Q-BEC

unconventional symm. beyond the “no-node” theorem.

p-wave BEC, and topological spin textures

• Fermions: Q-analytic Landau levels in 3D.

3

harmonic potential+ spin-orbit couplingCauchy-Riemann-Fueter condition.

• Complex quaternions: 3D Landau levels of Dirac fermions.

033 xx

History: how did people accept “i”?

• Not because of the “ridiculous”Eq. , but solving the cubic Eq (Cardano formula).

32

32

pq

3

2

23

2

13,2211 ,ii

ececxccx

32,1

2

qc

03 qpxx

• Start up with real coefficients, and end up with three real roots, but no way to avoid “i”.

4

12 x

!!!1,1

discriminant:

3,0 3,21 xx

The beauty of complex and elegancy

• 2D rotation: Gauss plane.

• Euler formula:

• Complex analysis based on complex analyticity:

• Applications: algebra fundamental theorem; Riemann hypothesis – distributions of prime numbers.

1ie

Quantum mechanics: the most important quantity in SchroedingerEq is not hbar but “i”.

Ht

i

5

0

y

gi

x

g)()(

1

2

10

0

zgzgdzzzi

Quaternion: a further extension

• Three imaginary units i, j, k.

ukzjyixq 1222 kji

• Hamilton: Non-commutative division algebra (3D rotation is non-commutative).

• Division algebra:

jikkiikjjkkjiij ;;

• Q-analyticity (Cauchy- Futer integral)

0

u

fk

z

fj

y

fi

x

f)()(

)(||

1

2

10

02

02

qfqfDqqqqq

0,00 baab

Quaternion plaque: Hamilton 10/16/1843

1222 ijkkjiBrougham bridge, Dublin

7

3D rotation as 1st Hopf map

• : imaginary quaternion unit:

3D rotation: unit quaternion q:2D rotation: unit complex phase

cossinsincossin)ˆ( kji

3

2sin)ˆ(

2cos Sq

• 3D rotation and Hopf map.

8

• 3D vector r imaginary quaternion. zkyjxir

• Rotation axis , rotation angle: .

21 Sqkq 3Sq

1st Hopf map

1

ˆ

qkqr

kzr

Outline

• Real numbers complex numbers quaternions (Q).

• Bosons: real (positive) BEC complex BEC Q-BEC

unconventional symm. beyond the “no-node” theorem.

p-wave BEC, and topological spin textures

• Fermions: Q-analytic Landau levels in 3D.

9

harmonic potential+ spin-orbit couplingCauchy-Riemann-Fueter condition.

• Complex quaternions: 3D Landau levels of Dirac fermions.

• “No-node” theorem: many-body ground state wavefunctions of bosons in the coordinate representation are positive-definite .

10

0),...,(21

n

rrr

N

jiji

N

iiex

iN

i

rrVrVM

H )()(2

int1

22

1

• A ground state property valid under general conditions (no rotation).

R. P. Feynman

Conventional BECs based on positive numbers

• No-go: Conventional BEC CANNOT break time-reversal symmspontaneously.

An intuitive proof

0),...,(21

n

rrr),...,( 21 nrrr |),...,(| 21 nrrr

)(|),..(|

)(|),..(||),...(|2

...||

int

2

1

1

2

1

2

11

2

1

jiji

n

n

iiexnni

n

in

rrVrr

rUrrrrm

drdrH

11

• More formally and rigorously, c.f. Perron-Frobenius theorem.

Conventional BEC (invariant under rotations, s-wave-like)

1212

• “No-node” theorem forbids unconventional symmetries, say, p, d, etc.

• Cf. conventional superconductivity: the pair WF belongs to the trivial (s-wave) representation of the rotation group.

)()()(

21

21 kekdrrrrik

k

k

• High Tc d-wave superconductivity tested by phase sensitive Josephson junction experiment by Van Harlingen et al.

Unconventional BEC: beyond “no-node”

• The condensate wavefunction belongs to a non-trivial (non-s-wave) representation of the lattice point group.

)(r

• No-node theorem does NOT apply to excited states.

• Example: the p-orbital bands can have degenerate band minima.

Interaction selects complex p+ip UBEC:

W. V. Liu, and C. Wu, PRA, 2006; Zi Cai, C. Wu, PRA 84, 033635(2011).Review: C. Wu, Mod. Phys. Lett. 23, 1(2009). Wirth, Olschlager, Hemmerich, Nat. Phys. 2011, N. Y. Kim, et al, Nat. Phy. Physics 2010.

13

)()()(21

rirrKK

K1 K2

-K1-K2

• Solid state experiments: d-wave BEC in exiton-polarition lattices (Kim, Yamamoto, Wu)

Outline

• Real numbers complex numbers quaternions (Q).

• Bosons: real (positive) BEC complex BEC Q-BEC

unconventional symm. beyond the “no-node” theorem.

p-wave BEC, and topological spin textures

• Fermions: Q-analytic Landau levels in 3D.

14

harmonic potential+ spin-orbit couplingCauchy-Riemann-Fueter condition.

• Complex quaternions: 3D Landau levels of Dirac fermions.

Two-component spinor quaternion

)ˆ,(r

• A quaternion can also be understood as a pair of complex numbers just like complex number is a pair of real numbers.

15kj

ir

ImRe

ImRe)ˆ,(

• Spinor wavefunction quaternion wavefunction;Spin density distribution 1st Hopf-mapping.

)ˆ,(2

)ˆ,()ˆ,( rrrS

kkSjSiS zyx2

1

The 2D version: BEC with Rashba SO coupling

16

• Free space: Spin spiral stripe for density-only interactions. (“order from disorder” mechanism beyond G-P level.)

• Spin-textures in SO coupled excitoncondensates observed from photoluminescence by L. Butov.

C. Wu , I. Mondragon Shem, and X.F. Zhou, Chin. Phys. Lett. 28, 097102 (2011)

(arXiv:0809.3532). – We are grateful to Chinese Physics Letters for the

acceleration of publication.

Exciton: A.A. High et al., Nature 483, 584 (2012). A.A. High et al., arXiv:1103.0321.

Cold atoms: Y. J. Lin et al, Nature 2011.

• Solid state SO coupled boson system: excitons in semiconductors.

Trap: Prediction of spontaneous generation of “baby” skyrmion spin-texture and half-quantum vortex.

3D SO coupling in hyperfine states of 40K atoms

)(2

1

2

2222

3

irmm

H D

Artificial SO coupling from light-atom interaction

).8

177,0,0(),0,

2

3,

2

1(

),0,1,0(),0,2

3,

2

1(

43

21

kk

kk

.166.0

).ˆ)(ˆˆ(166.0

m

eIeeA zzyyxx

Yi Li, Xiangfa Zhou, C. Wu, PRB

85, 125122 (2012).

Anderson, Juzeliunas, Spielman

and Galitskil,.Phys. Rev. Lett.

108, 235301 (2012).

Spherical rotator subjected to a fundamental monople

• Low energy SO sphere with radius kso. Dimensionless SO coupling strength

,)()(22

2

122

2

1 kAiMrMrVkTTex

22

22

2

1)(

2rmi

mH

T

sok

22

kAkd

• Momentum space:

TsoTso lkm

lmk

,,

• Angular momentum quantization change to half-integer values.18

19

• LLL wavefunctions through SO coupled harmonics:

Yi Li, Xiangfa Zhou, C. Wu, PRB 85, 125122 (2012); Gosh et al, PRA 84,

053629 (2011).

E

j0

rn

1r

n

• Angular dispersion suppressed by strong SO coupling forming nearly flat Landau levels.

TrTjn njj

Ezr

)2

1(

2

)1(2,

TR invariant parity breaking 3D Landau-levels

The single-particle ground state (mixed s and p-partial wave)

)sin

cos)(

0

1)(()ˆ,( 10

2

2

1

2

2

isoso

l

r

jj erkijrkjer T

z

• Quaternion representation:

20

)sin)ˆ((cos|)(|

|)(|

ImReImRe)ˆ,(

)()ˆ(

r

er

kjir

r

argument imaginary unit

• Imaginary unit solid angle direction

cossinsincossin)ˆ( kji

• Along any , winds in (1, )-plane as r increases. at zero points of g(r).

BEC WF as a quaternionic defect

• Mapping from 3D real space to quaternionic phase space – 3D skrymion.

2|)(| rf

2|)(| rgfgr /)(tan

)sin)ˆ((cos|)(|)ˆ,( rr

3S

• Condensation WF based on single-particle GS.

nrn)(

)ˆ()()ˆ( re

cossinsincossin)ˆ( kji

Quaternionic defect with non-zero Hopf invariant

22

• 3D spin distribution: the horizontal cross-section z=0 is a baby (2D) skrymion.

xy-plane

)2cossin)(cos()(

cos

cossin

cos,sin

sin,cossinsin)(

)(

)(

222

2

rrS

rrS

rS

z

y

x

5.0/ Tlz

See the linking number

• The trajectory of spin with the same orientation form a closed loop (in our case, the loop may be cut by the boundary).

• For any two loops, they link each other.

23

)2cossin)(cos()( 222 rrSz

• For ; reduced to a straight line of the z-axis

For ; a circle in the equatorial plane

,0,// zS

2,

2ˆ//

zS

3D quaternion analogy to 2D Abrikosov vortex lattice

24

2.0/ Tlz

• As SO coupling goes strong, the condensate WF mixes different single particle states within the same LL driven by interaction.

• Rotational symmetry is broken forming texture lattices.

0/ Tlz

1,8.0,15 c

Outline

• Real numbers complex numbers quaternions.

• Positive BEC complex BEC quaternionic BEC

go beyond the “no-node” theorem

p-wave BEC, and topological spin textures

• Quaternionic analytic Landau levels in 3D.

25

harmonic potential+ spin-orbit couplingCauchy-Riemann-Fueter condition.

• Complex quaternions: 3D Landau levels of Dirac fermions.

26

3D SU(2) LLs

2D LLs (continuum, flat spectra and complex analyticity); IQHE & FQHE.

2D QAH, Chern Insulator, TI; Frac-TI (lattice)

(symm-like gauge) degenerate helical modes;math: quaternionic analyticity

Dirac fermion: flat band of zero modes from non-minimal coupling; novel anomaly?

(Landau-like gauge) 4DQHE: spatially separated (3+1)d chiral anomaly

Exp. realization: cold atom; semiconductors?

3D TI (lattice)Frac-TI (3D

Laughlin WF)?

3ˆ// eS

2e

1e

QM: (Final exam for grad students. )harmonic oscillator + spin-orbit coupling

0

xk

yk

u

zk

BEh

enju

2

0

0

2

3

ai

aiH LLDiracD

symm

3D quaternionic (SU(2)) Landau levels

.0,

,)2/(|| 22

miyxz

ez BlzmsymLLL

Return to Landau levels (a review of 2D)

• Simple, explicit and elegant.

xiklkyyLandau

LLL

xBx ee)2/())(( 22

0

Landau gauge: 1D harmonic oscillators with center-shift (kx dep.)spatial separation of chiral modes.

xBkly 2

0

.0,,)2/(|| 22

miyxzez Blzmsym

LLL

Symmetric gauge: analytic functions of complex variables (chiral).

y

x

0

0xk

0xk

Comparison of symm gauge LLs in 2D and 3D

28

• 3D LLs: SU(2) group space

quaternionic analytic polynomials.

2

2

3

2

ˆ21,])ˆˆ[(),( gl

r

e

l

high

LLL

jereier

• 2D LLs: complex analytic polynomials.

.0,,)2/(|| 22

miyxzez Blzmsym

LLL Phase

Right-handed triad

• 1D harmonic levels: real polynomials.

3ˆ// eS

2e

1e

Conclusions

29

• Quaternion is a beautiful and useful concept.

• The 3D spin-orbit coupled BEC exhibit the 3D structure of quaternionic defects with non-zero Hopf invariants.

• Landau levels are generalized to 3D with the full rotational symmetry and TR symmetry.

• Quanternionic analyticity is a useful criterion to select good 3D wavefunctions for non-trivial topology.

Proposal: Quantum Mechanics class instruction (Final or qualexams).

k

RashbaE

sok

Landau-level (LL) type single particle quantization (2D)

• Ring rotor in momentum space subjected to a flux half integer quantized angular momentum.

,)()(22

2

122

2

1 kAiMrMrVkTTex

• The Rashba ring in momentum space with the radius kso. Dimensionless SO coupling strength .

1,/ TsoTTlkml

C. Wu et al arXiv:0809.3532; C. Wu et al,

Chin. Phys. Lett. 28, 097102 (2011). Hui Hu,

et al PRL 2012. 30

• Angular dispersion is suppressed by strong SO coupling forming nearly flat Landau levels.

const 2 2

2

,

T

z

Trjn

jnE

zr

dkkA )(

SO coupled bosons in trap: spin textures and half-quantum vortex

31

• Weak SO coupling. Condensate WFs:

,)(

)(2/1

ierg

rf

• Friedel-like oscillation of (up), and (down) at the pitch value of kso.

• The relative phase shift between f(r) and g(r).

2|)(| rf2|)(| rg

2|)(| rf

2|)(| rg

2

• 2D (baby) skyrmion type spin texture formation and half-quantum vortex.

sin,cos,|||| 22 fgSfgSgfSyxz

31

2/1 SLjz

)(rS

Spin lines with non-zero Hopf invariant

32

• 3D spin configuration with non-trivial Hopf-invariant; twisted baby (2D) skrymion as moving along z-axis.