1 of 16

TUNING TOPOLOGICAL PROPERTIES: ONE STEP TOWARDS 2D QUANTUM COMPUTING

Karunya Shirali, William Shelton and Ilya Vekhter

Louisiana State University

2 of 16

MOTIVATION

  • Topological insulators (TIs): �insulators in the bulk, topologically protected metallic surface states
  • Topological quantum computing:�information encoded in braiding of anyons
  • Class of systems predicted to host anyons:�Topological superconductors, etc

  • Hard to observe experimentally
  • Alternative: TI-superconductor heterostructure�Proximity induced superconductivity

��resistant to decoherence

Topological insulator

 

Surface states[1]

3 of 16

SPIN-ORBIT COUPLING AND HELICAL SURFACE STATES

  •  

Helical state2 (k space)

Helical state1 (real space)

Surface states observed with ARPES3

4 of 16

LEVEL SPLITTING AND SPIN-ORBIT COUPLING

  • Bi2Se3: Start from atomic p orbitals1 of Bi, Se

  • Consider chemistry of the system

  • Find Bi pz, Se pz levels near chemical potential

  • Include SOC:�Obtain band inversion �

1: PRB 82, 045122 (2010)

Band inversion

(i) (ii) (iii) (iv)

5 of 16

WHY DO WE USE DFT?

  •  

6 of 16

BI2SE3 AND BI2TE3: EXISTING WORK AND PROBLEMS

  • Most calculations keep lattice parameters fixed without optimizing them
  • Reason: some lattice parameters, especially dint, are significantly different from experiment if they aren’t fixed�
  • But DFT is a ground-state theory: could obtain metastable states�
  • We fill this gap and set up systematic means1 of performing first-principles calculations on 3D TIs
  • Obtain bulk lattice constants, inter-QL distances,�properties of topological surface states accurately�
  • Important to include SOC�� 1: K Shirali, W A Shelton and I Vekhter 2021 Journal of Physics: Condensed Matter 33 035702

7 of 16

REPRODUCE BULK, SURFACE PROPERTIES

  •  

Hexagonal warping

8 of 16

METHODOLOGY TO INCLUDE VDW FOR TIS

  • Set up systematic way1 to calculate structural & electronic properties self-consistently
  • Bulk: lattice parameters and band gap, inter-QL distance�Surface: Dirac velocities and spin textures �
  • Have methodology for: bulk ✅� surface ✅

  • Moving forward: how does strain affect topological properties?�

1: Shirali K, Shelton W A and Vekhter I 2020 Journal of Physics: Condensed Matter 33 035702

9 of 16

TOPOLOGY FROM BAND INVERSION

  • Bi2Se3 under strain, fix c/a, Phydr = 0
  • QL remains intact; inter-QL distance expands as c increases

Inverted bands

No inversion

 

Bi2Se3 bulk bands

10 of 16

COULOMB AND VDW DRIVE TRANSITION

  •  

Figures:

(a) Bi2Se3 band gap

(b) Bi pz, Se pz levels

as a fn. of c/a

Se1 pz (outer atomic layers of QL) levels change with c/a:

Mainly inter-QL physics

Transition

Level switching

at transition

11 of 16

COULOMB AND VDW DRIVE TRANSITION

  •  

Figures:

(a) Bi2Se3 band gap

(b) Bi pz, Se pz levels

as a fn. of c/a

Se1 pz (outer atomic layers of QL) levels change with c/a:

Mainly inter-QL physics

Transition

Topological

Trivial

12 of 16

COULOMB VERSUS VDW

  • Emergent picture: Se-Se (Te-Te) Coulomb repulsion �between QLs compete with vdW attraction �as c/a is varied �Drives transition
  • Subtract ionic radii:�universal value for critical atomic spacing

Can be measured in experiment

arXiV 2202.12867

13 of 16

TOPOLOGICAL PHASE TRANSITIONS IN DISORDERED TIS

  • Topological phase transition in (Bi1-xSbx)2Se3: Sb substitution reduces strength of SOC
  • Range of values (78%-83%) predicted1 for critical impurity concentration
  • Do Bi and Sb contribute to the same bands or distinct bands?

  • Simulating disorder in DFT: large unit cells required to mimic randomness
  • SCRAPS2 generate supercells for Bi/Sb alloying, atomic pair correlations �zero up to third nearest neighbor: no configurational averaging required�

In collaboration with Ames Laboratory

1: Liu J and Vanderbilt D 2013 Phys. Rev. B 88 224202

2: Nature Computational Science 1, 54–61 (2021)

Figure: (Bi1-xSbx)2Se3 crystal structure, x = 0.5

dint

Sb

Se

Bi

QL

14 of 16

CONCLUDING REMARKS, OUTLOOK

  • 50%, 62.5%, 75% concentrations: obtain trivial insulator ground-states from SCRAPS supercells
  • But end-point materials both topological: �Two topological transitions, one at low x and other �at large x

Future work

  • Bi2(Se1-yTey)3 alloy: �- Se and Te atoms have two sublattices to choose from

x = 0.5, 0.625, 0.75

x = 0.02

15 of 16

CONCLUSION

  • Set up comprehensive framework for ab initio calculations of topological insulators�
  • Reproduce bulk, surface properties accurately�
  • Explained essential physics of topological phase transitions under strain�Due to competing Coulomb and van der Waals interactions between the quintuple layers �
  • Foundation to study and address other problems involving topological insulators

16 of 16

ACKNOWLEDGEMENTS

This research was supported by NSF via Grant No. DMR-1410741 and by the U.S. Department of Energy under EPSCoR Grant No. DE-SC0012432 with additional support from the Louisiana Board of Regents. Portions of this research were conducted with high performance computing resources provided by Louisiana State University (http://www.hpc.lsu.edu).

Thank you!