simulating few- and many-body physics with rydberg...
TRANSCRIPT
FORTHIESL
FORTHIESL
Simulating few- and many-body physics
with Rydberg atoms
David Petrosyan
Heraklion, 25/10/17 – p. 1/29
FORTHIESL
FORTHIESLPlan of the Seminar
Quantum Simulations and Computations
Heraklion, 25/10/17 – p. 2/29
FORTHIESL
FORTHIESLPlan of the Seminar
Quantum Simulations and Computations
Rydberg atoms and their interactions
Heraklion, 25/10/17 – p. 2/29
FORTHIESL
FORTHIESLPlan of the Seminar
Quantum Simulations and Computations
Rydberg atoms and their interactions
Quantum gates
Heraklion, 25/10/17 – p. 2/29
FORTHIESL
FORTHIESLPlan of the Seminar
Quantum Simulations and Computations
Rydberg atoms and their interactions
Quantum gates
Simulating lattice spin models in 1D & 2D
Heraklion, 25/10/17 – p. 2/29
FORTHIESL
FORTHIESLPlan of the Seminar
Quantum Simulations and Computations
Rydberg atoms and their interactions
Quantum gates
Simulating lattice spin models in 1D & 2D
Summary
Heraklion, 25/10/17 – p. 2/29
FORTHIESL
FORTHIESL
Outline of Quantum Theory
Heraklion, 25/10/17 – p. 3/29
FORTHIESL
FORTHIESL
Quantum state in a Hilbert Space
|Ψ〉 is a vector in a complex vector space H
|Ψ〉 = c1 |e1〉+ c2 |e2〉+ c3 |e3〉+ . . . =∑N
i=1 ci |ei〉∑N
i=1 |ci|2 = 1
basis |ei〉: eigenstates of some operator O |ei〉 = Oi |ei〉
Heraklion, 25/10/17 – p. 4/29
FORTHIESL
FORTHIESL
Quantum state in a Hilbert Space
|Ψ〉 is a vector in a complex vector space H
|Ψ〉 = c1 |e1〉+ c2 |e2〉+ c3 |e3〉+ . . . =∑N
i=1 ci |ei〉∑N
i=1 |ci|2 = 1
basis |ei〉: eigenstates of some operator O |ei〉 = Oi |ei〉
Composite system S = A+ B + . . .:
dimHS = dimHA × dimHB × . . .
Heraklion, 25/10/17 – p. 4/29
FORTHIESL
FORTHIESL
Quantum state in a Hilbert Space
|Ψ〉 is a vector in a complex vector space H
|Ψ〉 = c1 |e1〉+ c2 |e2〉+ c3 |e3〉+ . . . =∑N
i=1 ci |ei〉∑N
i=1 |ci|2 = 1
basis |ei〉: eigenstates of some operator O |ei〉 = Oi |ei〉
Composite system S = A+ B + . . .:
dimHS = dimHA × dimHB × . . .
Two state system: spin- 12 electron | ↑〉, | ↓〉
Ψz
yx
Bloch Sphere
Heraklion, 25/10/17 – p. 4/29
FORTHIESL
FORTHIESL
Quantum state in a Hilbert Space
|Ψ〉 is a vector in a complex vector space H
|Ψ〉 = c1 |e1〉+ c2 |e2〉+ c3 |e3〉+ . . . =∑N
i=1 ci |ei〉∑N
i=1 |ci|2 = 1
basis |ei〉: eigenstates of some operator O |ei〉 = Oi |ei〉
Composite system S = A+ B + . . .:
dimHS = dimHA × dimHB × . . .
Two state system: spin- 12 electron | ↑〉, | ↓〉
Ψz
yx
Bloch Sphere⇒ For N two-state systems: dimHS = 2N
Heraklion, 25/10/17 – p. 4/29
FORTHIESL
FORTHIESL
Quantum state in a Hilbert Space
|Ψ〉 is a vector in a complex vector space H
|Ψ〉 = c1 |e1〉+ c2 |e2〉+ c3 |e3〉+ . . . =∑N
i=1 ci |ei〉∑N
i=1 |ci|2 = 1
basis |ei〉: eigenstates of some operator O |ei〉 = Oi |ei〉
Composite system S = A+ B + . . .:
dimHS = dimHA × dimHB × . . .
Two state system: spin- 12 electron | ↑〉, | ↓〉
Ψz
yx
Bloch Sphere⇒ For N two-state systems: dimHS = 2N
Hilbert Space is a big place!
(Carlton M. Caves)
Heraklion, 25/10/17 – p. 4/29
FORTHIESL
FORTHIESLTime evolution
Hamiltonian operator H: Energy 〈Ψ| H |Ψ〉 = E
Heraklion, 25/10/17 – p. 5/29
FORTHIESL
FORTHIESLTime evolution
Hamiltonian operator H: Energy 〈Ψ| H |Ψ〉 = E
Our favorite operator!
Heraklion, 25/10/17 – p. 5/29
FORTHIESL
FORTHIESLTime evolution
Hamiltonian operator H: Energy 〈Ψ| H |Ψ〉 = E
Our favorite operator!
Schrödinger equation i~ ∂∂t
|Ψ〉 = H |Ψ〉 ⇒ |Ψ(t)〉 = e−i~Ht |Ψ(0)〉
Heraklion, 25/10/17 – p. 5/29
FORTHIESL
FORTHIESLTime evolution
Hamiltonian operator H: Energy 〈Ψ| H |Ψ〉 = E
Our favorite operator!
Schrödinger equation i~ ∂∂t
|Ψ〉 = H |Ψ〉 ⇒ |Ψ(t)〉 = e−i~Ht |Ψ(0)〉
Mixed states, dissipative systems: ρ =∑
Ψ PΨ |Ψ〉〈Ψ|
Liouville – von Neumann equation i~ ∂∂tρ = [Hρ− ρH] + Lρ
Heraklion, 25/10/17 – p. 5/29
FORTHIESL
FORTHIESL
Quantum simulations
Richard Feynman
Simulating Physics with Computers,Feynman, Int. J. Theor. Phys. 21, 467 (1982);
Lloyd, Science 273, 1073 (1996)
Heraklion, 25/10/17 – p. 6/29
FORTHIESL
FORTHIESL
Quantum simulations
Richard Feynman
Simulating Physics with Computers,Feynman, Int. J. Theor. Phys. 21, 467 (1982);
Lloyd, Science 273, 1073 (1996)
Interacting many-body quantum systemsare hard to simulate on classical computers
Many degrees of freedom (huge HS)
Quantum correlations (entanglement)
|Ψ〉S 6= |Ψ〉A ⊗ |Ψ〉B ⊗ . . .
Heraklion, 25/10/17 – p. 6/29
FORTHIESL
FORTHIESL
Quantum simulations
Richard Feynman
Simulating Physics with Computers,Feynman, Int. J. Theor. Phys. 21, 467 (1982);
Lloyd, Science 273, 1073 (1996)
Interacting many-body quantum systemsare hard to simulate on classical computers
Many degrees of freedom (huge HS)
Quantum correlations (entanglement)
|Ψ〉S 6= |Ψ〉A ⊗ |Ψ〉B ⊗ . . .
⇒ number of c equations to solve simultaneously is dimHS
decoupling (mean field) approximations are not always correct
Heraklion, 25/10/17 – p. 6/29
FORTHIESL
FORTHIESL
Quantum simulations
Richard Feynman
Simulating Physics with Computers,Feynman, Int. J. Theor. Phys. 21, 467 (1982);
Lloyd, Science 273, 1073 (1996)
Interacting many-body quantum systemsare hard to simulate on classical computers
Many degrees of freedom (huge HS)
Quantum correlations (entanglement)
|Ψ〉S 6= |Ψ〉A ⊗ |Ψ〉B ⊗ . . .
⇒ number of c equations to solve simultaneously is dimHS
decoupling (mean field) approximations are not always correct
Universal quantum simulator
Nat. Phys. 6, 382 (2010)
Discretize space & time [ST eδ(A+B) = eδAeδB +O(δ2)]
⇒ spin lattice with finite range & interval interactions
U(t) = exp(− i~Ht) ≃ ∏
j exp(− i~Hjδt1)
∏
j′ exp(− i~Hj′δt2) . . .
Heraklion, 25/10/17 – p. 6/29
FORTHIESL
FORTHIESL
Quantum simulations
Richard Feynman
Simulating Physics with Computers,Feynman, Int. J. Theor. Phys. 21, 467 (1982);
Lloyd, Science 273, 1073 (1996)
Interacting many-body quantum systemsare hard to simulate on classical computers
Many degrees of freedom (huge HS)
Quantum correlations (entanglement)
|Ψ〉S 6= |Ψ〉A ⊗ |Ψ〉B ⊗ . . .
⇒ number of c equations to solve simultaneously is dimHS
decoupling (mean field) approximations are not always correct
Analog quantum simulator
Nature 415, 39 (2002)
Construct a clean system &realize appropriate interactions to mimic HS
⇒ Dynamically or adiabatically evolve |Ψ〉and read-out
Heraklion, 25/10/17 – p. 6/29
FORTHIESL
FORTHIESL
Quantum computations
0
1
Quantum bit – qubit – is two-state quantum system
stores superposition states |ψ〉 = α |0〉+ β |1〉
Heraklion, 25/10/17 – p. 7/29
FORTHIESL
FORTHIESL
Quantum computations
0
1
Quantum bit – qubit – is two-state quantum system
stores superposition states |ψ〉 = α |0〉+ β |1〉
N -qubit memory has 2N orthogonal states |x〉 = |00 . . . 0〉, . . . , |11 . . . 1〉|Ψ〉 = ∑
x cx |x〉 in general |Ψ〉 6= |ψ〉1 ⊗ |ψ〉2 ⊗ . . .: entanglement
Heraklion, 25/10/17 – p. 7/29
FORTHIESL
FORTHIESL
Quantum computations
0
1
Quantum bit – qubit – is two-state quantum system
stores superposition states |ψ〉 = α |0〉+ β |1〉
N -qubit memory has 2N orthogonal states |x〉 = |00 . . . 0〉, . . . , |11 . . . 1〉|Ψ〉 = ∑
x cx |x〉 in general |Ψ〉 6= |ψ〉1 ⊗ |ψ〉2 ⊗ . . .: entanglement
Any computation (unitary transformation) can be realizedvia a sequence of one-, two- (and more-) qubit gates
One-qubit gates (rotations on BS): Unity I, Hadamard H, Pauli X,Y, Z, Phase S ...
Two-qubit gates: Controlled-U , Controlled-Z, CNOT= HtCZHt, SWAP,√
SWAP ...
Heraklion, 25/10/17 – p. 7/29
FORTHIESL
FORTHIESL
Quantum computations
0
1
Quantum bit – qubit – is two-state quantum system
stores superposition states |ψ〉 = α |0〉+ β |1〉
N -qubit memory has 2N orthogonal states |x〉 = |00 . . . 0〉, . . . , |11 . . . 1〉|Ψ〉 = ∑
x cx |x〉 in general |Ψ〉 6= |ψ〉1 ⊗ |ψ〉2 ⊗ . . .: entanglement
Any computation (unitary transformation) can be realizedvia a sequence of one-, two- (and more-) qubit gates
One-qubit gates (rotations on BS): Unity I, Hadamard H, Pauli X,Y, Z, Phase S ...
Two-qubit gates: Controlled-U , Controlled-Z, CNOT= HtCZHt, SWAP,√
SWAP ...
Quantum computation is inherently parallel
Prepare the input state in a superposition state of all possible “classical” inputs x:
|Ψinput〉 =∑
x cx |x〉Design an algorithm where all the computational paths interfere with each other to
yield with high probability the output state y: |Ψoutput〉 ≃ |y〉 (∑
y′ 6=y |cy′ |2 → 0)
Heraklion, 25/10/17 – p. 7/29
FORTHIESL
FORTHIESL
Quantum computations
0
1
Quantum bit – qubit – is two-state quantum system
stores superposition states |ψ〉 = α |0〉+ β |1〉
N -qubit memory has 2N orthogonal states |x〉 = |00 . . . 0〉, . . . , |11 . . . 1〉|Ψ〉 = ∑
x cx |x〉 in general |Ψ〉 6= |ψ〉1 ⊗ |ψ〉2 ⊗ . . .: entanglement
Any computation (unitary transformation) can be realizedvia a sequence of one-, two- (and more-) qubit gates
One-qubit gates (rotations on BS): Unity I, Hadamard H, Pauli X,Y, Z, Phase S ...
Two-qubit gates: Controlled-U , Controlled-Z, CNOT= HtCZHt, SWAP,√
SWAP ...
Quantum computation is inherently parallel
Prepare the input state in a superposition state of all possible “classical” inputs x:
|Ψinput〉 =∑
x cx |x〉Design an algorithm where all the computational paths interfere with each other to
yield with high probability the output state y: |Ψoutput〉 ≃ |y〉 (∑
y′ 6=y |cy′ |2 → 0)
Useful quantum algorithms:
Grover search of M = 2N elements in√M steps
Shor FFT for a set M = 2N in ∼ N2 steps (factorization etc.) ...Heraklion, 25/10/17 – p. 7/29
FORTHIESL
FORTHIESL
Rydberg atoms
Niels Henrik David Bohr
Heraklion, 25/10/17 – p. 8/29
FORTHIESL
FORTHIESL
Rydberg atoms
Niels Henrik David Bohr
Heraklion, 25/10/17 – p. 8/29
FORTHIESL
FORTHIESL
Rydberg Atoms
High principal quantum number
n ≫ 1 (H-like)
Heraklion, 25/10/17 – p. 9/29
FORTHIESL
FORTHIESL
Rydberg Atoms
High principal quantum number
n ≫ 1 (H-like)r
g
Energy Er = − Ryn∗2
effective PQN n∗ = n− δl (δl quantum defect)
Heraklion, 25/10/17 – p. 9/29
FORTHIESL
FORTHIESL
Rydberg Atoms
High principal quantum number
n ≫ 1 (H-like)r
g
Energy Er = − Ryn∗2
effective PQN n∗ = n− δl (δl quantum defect)
0a2n+ −
Easily polarized
Huge dipole moments ℘ ∼ n2ea0
Gallagher, Rydberg Atoms (Cambridge 1994)Heraklion, 25/10/17 – p. 9/29
FORTHIESL
FORTHIESL
Dipole-Dipole Interactions
D =℘(1) · ℘(2)
R3− 3
(℘(1) ·R)(℘(2) ·R)
R5∝ n4
Heraklion, 25/10/17 – p. 10/29
FORTHIESL
FORTHIESL
Dipole-Dipole Interactions
D =℘(1) · ℘(2)
R3− 3
(℘(1) ·R)(℘(2) ·R)
R5∝ n4
⇒ Static DDI
Est induced Stark eigenstates
with permanent ℘ = 32nqea0
q
−q
Est
−q+1
q−1
−
+
Heraklion, 25/10/17 – p. 10/29
FORTHIESL
FORTHIESL
Dipole-Dipole Interactions
D =℘(1) · ℘(2)
R3− 3
(℘(1) ·R)(℘(2) ·R)
R5∝ n4
⇒ Static DDI
Est induced Stark eigenstates
with permanent ℘ = 32nqea0
pθ
(2)p(1)
RD = ℘(1)℘(2)(1−3 cos2 θ)
R3
Heraklion, 25/10/17 – p. 10/29
FORTHIESL
FORTHIESL
Dipole-Dipole Interactions
D =℘(1) · ℘(2)
R3− 3
(℘(1) ·R)(℘(2) ·R)
R5∝ n4
⇒ Resonant DDI
D12
D121
2
2a
r
b
r
1
D = 1R3
[
℘(1)+1℘
(2)−1 + ℘
(1)−1℘
(2)+1 + ℘
(1)0 ℘
(2)0 (1− 3 cos2 θ)
− 32sin2 θ(℘
(1)+1℘
(2)+1 + ℘
(1)+1℘
(2)−1 + ℘
(1)−1℘
(2)+1 + ℘
(1)−1℘
(2)−1
− 3√2sin θ cos θ(℘
(1)+1℘
(2)0 + ℘
(1)−1℘
(2)0 + ℘
(1)0 ℘
(2)+1 + ℘
(1)0 ℘
(2)−1)
]
Heraklion, 25/10/17 – p. 11/29
FORTHIESL
FORTHIESL
Dipole-Dipole Interactions
D =℘(1) · ℘(2)
R3− 3
(℘(1) ·R)(℘(2) ·R)
R5∝ n4
⇒ Resonant DDI
D12
D121
2
2a
r
b
r
1
D = 1R3
[
℘(1)+1℘
(2)−1 + ℘
(1)−1℘
(2)+1 + ℘
(1)0 ℘
(2)0 (1− 3 cos2 θ)
− 32sin2 θ(℘
(1)+1℘
(2)+1 + ℘
(1)+1℘
(2)−1 + ℘
(1)−1℘
(2)+1 + ℘
(1)−1℘
(2)−1
− 3√2sin θ cos θ(℘
(1)+1℘
(2)0 + ℘
(1)−1℘
(2)0 + ℘
(1)0 ℘
(2)+1 + ℘
(1)0 ℘
(2)−1)
]
℘+1 = − e√2(x+ iy) = − e√
2r sin θeiφ = er
√
4π3Y1,1(θ, φ) r+1 = 〈n′l′,m+ 1| r |nl,m〉
℘0 = ez = er cos θ = er√
4π3Y1,0(θ, φ) r0 = 〈n′l′,m| r |nl,m〉
℘−1 = e√2(x− iy) = e√
2r sin θe−iφ = er
√
4π3Y1,−1(θ, φ) r−1 = 〈n′l′,m− 1| r |nl,m〉
Heraklion, 25/10/17 – p. 11/29
FORTHIESL
FORTHIESLvan der Waals Interaction
12D
D12r
aatom
a1atom
r
δ δ
D21
D21
bb
2RDDI (Förster process)
D12 ≡ D(R) ∝ ℘br℘ar
R3 ∝ n4
Heraklion, 25/10/17 – p. 12/29
FORTHIESL
FORTHIESLvan der Waals Interaction
12D
D12r
aatom
a1atom
r
δ δ
D21
D21
bb
2RDDI (Förster process)
D12 ≡ D(R) ∝ ℘br℘ar
R3 ∝ n4
ωrb − ωar = δω ≫ D
(δω ∝ n−3)
⇒ |r1〉 |r2〉 9 |a1,2〉 |b2,1〉: Non-Resonant DDI (Adiabatic elim. |a1,2〉 |b2,1〉)
Heraklion, 25/10/17 – p. 12/29
FORTHIESL
FORTHIESLvan der Waals Interaction
r
aatom 2
a1
r
bb
atom
12∆
RDDI (Förster process)
D12 ≡ D(R) ∝ ℘br℘ar
R3 ∝ n4
ωrb − ωar = δω ≫ D
(δω ∝ n−3)
⇒ |r1〉 |r2〉 9 |a1,2〉 |b2,1〉: Non-Resonant DDI (Adiabatic elim. |a1,2〉 |b2,1〉)
⇒ Energy shift of |r1〉 |r2〉 (2nd-order in D/δω)
R
∆
VvdW = ~σ1rr∆12σ
2rr
∆12 ≡ ∆(R) = 2|D(R)|2
δω= C6
R6 ∝ n11 — vdWI strength
Saffman, Walker, Mølmer, RMP 82, 2313 (2010)
Heraklion, 25/10/17 – p. 12/29
FORTHIESL
FORTHIESL
Two Atoms
Heraklion, 25/10/17 – p. 13/29
FORTHIESL
FORTHIESL
Rydberg interaction blockade
2Ω
2Ω
2rg
gg
rr
+gr
∆Two-atom Hamiltonian
H/~ = Ω |r〉1〈g| +Ω |r〉2〈g| +H.c.
+∆12 |r〉1〈r| ⊗ |r〉2〈r|
Heraklion, 25/10/17 – p. 14/29
FORTHIESL
FORTHIESL
Rydberg interaction blockade
2Ω
2Ω
2rg
gg
rr
+gr
∆Two-atom Hamiltonian
H/~ = Ω |r〉1〈g| +Ω |r〉2〈g| +H.c.
+∆12 |r〉1〈r| ⊗ |r〉2〈r|
Resonant transition |gg〉 ↔ |gr, rg〉1~〈gr, rg|H |gg〉 = Ω
Nonresonant transition |gr, rg〉 ↔ |rr〉1~〈rr| H |rr〉 = ∆12 ≫ Ω
Heraklion, 25/10/17 – p. 14/29
FORTHIESL
FORTHIESL
Rydberg interaction blockade
2Ω
2Ω
2rg
gg
rr
+gr
∆Two-atom Hamiltonian
H/~ = Ω |r〉1〈g| +Ω |r〉2〈g| +H.c.
+∆12 |r〉1〈r| ⊗ |r〉2〈r|
Resonant transition |gg〉 ↔ |gr, rg〉1~〈gr, rg|H |gg〉 = Ω
Nonresonant transition |gr, rg〉 ↔ |rr〉1~〈rr| H |rr〉 = ∆12 ≫ Ω
⇒ Double excitation |rr〉 is blocked
⇒ Rabi oscillations |gg〉 ↔ 1√2( |gr〉+ |rg〉) with
√2Ω
Jaksch et al., PRL 85, 2208 (2000)Heraklion, 25/10/17 – p. 14/29
FORTHIESL
FORTHIESLTwo-atom Rabi oscillations
a
b
R = 18 µm
R = 3.6 µm
Duration of the excitation (ns)
Excita
tio
n p
rob
ab
ility
E
xcita
tio
n p
rob
ab
ility
0 40 80 120 160 200 240 280 320
0.0
1.0
0.2
0.4
0.6
0.8
0.0
0.2
0.4
0.6
0.8
Urban et al., Nature Phys. 5, 110 (2009); Gaëtan et al., Nature Phys. 5, 115 (2009)
Heraklion, 25/10/17 – p. 15/29
FORTHIESL
FORTHIESLTwo-atom Rabi oscillations
a
b
R = 18 µm
R = 3.6 µm
Duration of the excitation (ns)
Excita
tio
n p
rob
ab
ility
E
xcita
tio
n p
rob
ab
ility
0 40 80 120 160 200 240 280 320
0.0
1.0
0.2
0.4
0.6
0.8
0.0
0.2
0.4
0.6
0.8
Urban et al., Nature Phys. 5, 110 (2009); Gaëtan et al., Nature Phys. 5, 115 (2009)
Duration of the excitation (ns)
Excita
tio
n p
rob
ab
ility
0 40 80 120 160 200 240 280 320
0.0
1.0
0.2
0.4
0.6
0.8 R = 3.6 µm
Heraklion, 25/10/17 – p. 15/29
FORTHIESL
FORTHIESL
Rydberg blockade gates
r
2Ω1Ω
Atom 2
2
∆12?r
Atom 1
1 3
0110
Two-atom (qubit) CZ gate
|0〉1 |0〉2 → |0〉1 |0〉2|0〉1 |1〉2 → −|0〉1 |1〉2|1〉1 |0〉2 → −|1〉1 |0〉2|1〉1 |1〉2 → −|1〉1 |1〉2
Jaksch et al., PRL 85, 2208 (2000)
Heraklion, 25/10/17 – p. 16/29
FORTHIESL
FORTHIESL
Rydberg blockade gates
r
2Ω1Ω
Atom 2
2
∆12?r
Atom 1
1 3
1 0 0 1
Two-atom (qubit) CNOT gate
|0〉1 |0〉2 → |0〉1 |0〉2|0〉1 |1〉2 → |0〉1 |1〉2|1〉1 |0〉2 → |1〉1 |1〉2|1〉1 |1〉2 → |1〉1 |0〉2
Heraklion, 25/10/17 – p. 16/29
FORTHIESL
FORTHIESL
Rydberg blockade gates
r
2Ω1Ω
Atom 2
2
∆12?r
Atom 1
1 3
1 0 0 1
Two-atom (qubit) CNOT gate
|0〉1 |0〉2 → |0〉1 |0〉2|0〉1 |1〉2 → |0〉1 |1〉2|1〉1 |0〉2 → |1〉1 |1〉2|1〉1 |1〉2 → |1〉1 |0〉2
Wilk et al., PRL 104, 010502 (2010); Isenhower et al., PRL 104, 010503 (2010)Heraklion, 25/10/17 – p. 16/29
FORTHIESL
FORTHIESL
Spatially extended systems
Heraklion, 25/10/17 – p. 17/29
FORTHIESL
FORTHIESL
Lattice spin models
g
r
Ω
δ
Vaa
Heraklion, 25/10/17 – p. 18/29
FORTHIESL
FORTHIESL
Lattice spin models
g
r
Ω
δ
Vaa
spin- 12 Hamiltonian (Ising-like)
Hspin = ~∑N
j [Ωσjx − δj
2 σjz ] +
~
4
∑Ni 6=j ∆ij σ
izσ
jz
with Ω, δj = (δ − 12
∑
i 6=j ∆ij) & ∆ij ≡ ∆(xi − xj)
isotropic or anisotropic interaction
DD (∝ 1/R3) or vdW (∝ 1/R6)
Heraklion, 25/10/17 – p. 18/29
FORTHIESL
FORTHIESL
Lattice spin models
g
r
Ω
δ
Vaa
spin- 12 Hamiltonian (Ising-like)
Hspin = ~∑N
j [Ωσjx − δj
2 σjz ] +
~
4
∑Ni 6=j ∆ij σ
izσ
jz
with Ω, δj = (δ − 12
∑
i 6=j ∆ij) & ∆ij ≡ ∆(xi − xj)
isotropic or anisotropic interaction
DD (∝ 1/R3) or vdW (∝ 1/R6)
Phase diagram
(mean field)
δ
Weimer et al., PRL 101, 250601 (2008)
Heraklion, 25/10/17 – p. 18/29
FORTHIESL
FORTHIESL
Lattice spin models
g
r
Ω
δ
Vaa
spin- 12 Hamiltonian (Ising-like)
Hspin = ~∑N
j [Ωσjx − δj
2 σjz ] +
~
4
∑Ni 6=j ∆ij σ
izσ
jz
with Ω, δj = (δ − 12
∑
i 6=j ∆ij) & ∆ij ≡ ∆(xi − xj)
isotropic or anisotropic interaction
DD (∝ 1/R3) or vdW (∝ 1/R6)
Phase diagram
(mean field)
δ
Weimer et al., PRL 101, 250601 (2008)
Schachenmayer et al, NJP 12, 103044 (2010)Heraklion, 25/10/17 – p. 18/29
FORTHIESL
FORTHIESL
Dynamical crystal preparation (1D)
Ω → 0
H/~ = −δ∑N
j σjrr − Ω
∑Nj (σj
rg + σjgr) +
∑Ni<j ∆ij σ
irrσ
jrr
Heraklion, 25/10/17 – p. 19/29
FORTHIESL
FORTHIESL
Dynamical crystal preparation (1D)
Ω → 0
H/~ = −δ∑N
j σjrr − Ω
∑Nj (σj
rg + σjgr) +
∑Ni<j ∆ij σ
irrσ
jrr
Avoided crossings ±Ωnn+1 6= 0:
Ω01 =
√NΩ, Ω1
2 = 2Ω/√N, Ω2
3 = Ω, Ω34 ≃ Ω3/[C6/(l/3)
6]2 . . .
Pohl, Demler, Lukin, PRL 104, 043002 (2010)
Schachenmayer, Lesanovsky, Micheli, Daley, NJP 12, 103044 (2010)
Petrosyan, Mølmer, Fleischhauer, JPB 49, 084003 (2016)
Heraklion, 25/10/17 – p. 19/29
FORTHIESL
FORTHIESL
Dynamical crystal preparation (1D)
Experiment
Schauß et al., Science 347 1455 (2015); Bernien et al., arXiv:1707.04344Heraklion, 25/10/17 – p. 20/29
FORTHIESL
FORTHIESL
Simulations of adiabatic dynamics (QMC WF)
Pminn ≡
〈Rminn | ρ |Rmin
n 〉
0
0.1
0.2Ω
(2π
MH
z)
−1
0
1
δ (2
π M
Hz)
0 2 4 6 8 10 120
0.5
1
Pro
babi
litie
s
012345
Time τ
µs
Γr=Γz=0
1
Heraklion, 25/10/17 – p. 21/29
FORTHIESL
FORTHIESL
Simulations of adiabatic dynamics (QMC WF)
Pminn ≡
〈Rminn | ρ |Rmin
n 〉
0
0.1
0.2Ω
(2π
MH
z)
−1
0
1
δ (2
π M
Hz)
0
0.5
1
Pro
babi
litie
s
012345
0 2 4 6 8 10 120
0.5
1
Time τ
µs
Γr=Γz=0
1
1
Γr=40 kHzΓz=0
Heraklion, 25/10/17 – p. 21/29
FORTHIESL
FORTHIESL
Simulations of adiabatic dynamics (QMC WF)
Pminn ≡
〈Rminn | ρ |Rmin
n 〉
0
0.1
0.2Ω
(2π
MH
z)
−1
0
1
δ (2
π M
Hz)
0
0.5
1
Pro
babi
litie
s
012345
0
0.5
1
0 2 4 6 8 10 120
0.5
1
Time τ
µs
Γr=Γz=0
1
1
1
Γr=40 kHzΓz=0
Γz=0Γr=40 kHz
Heraklion, 25/10/17 – p. 21/29
FORTHIESL
FORTHIESL
Simulations of adiabatic dynamics (QMC WF)
Pminn ≡
〈Rminn | ρ |Rmin
n 〉
0
0.1
0.2Ω
(2π
MH
z)
−1
0
1
δ (2
π M
Hz)
0
0.5
1
Pro
babi
litie
s
012345
0
0.5
1
0
0.5
1
0 2 4 6 8 10 12Time
0
0.5
1
Time τ
µs
Γr=Γz=0
1
1
1
1
Γr=40 kHzΓz=0
Γz=0Γr=40 kHz
Γr=40 kHzΓz=40 kHz
Heraklion, 25/10/17 – p. 21/29
FORTHIESL
FORTHIESL
Simulations of adiabatic dynamics
F = Pmin3
4 8 12 16 20−1
−0.8
−0.6
0
0.2
0.4
0.6
0.8
1F
idel
ity
Γr=Γz=0 (kHz)Γr=40, Γz=0Γr=0, Γz=40Γr=Γz=40
0 4 8 12 16 20Preparation time τ (µs)
1
1.5
2
2.5
3
Mea
n ex
cita
tion
num
ber
<n>
4 8 12 16 200
0.5
1
1.5
Man
del Q
τ
τ
w
Heraklion, 25/10/17 – p. 22/29
FORTHIESL
FORTHIESL
Simulations of adiabatic dynamics
F = Pmin3
4 8 12 16 20−1
−0.8
−0.6
0
0.2
0.4
0.6
0.8
1F
idel
ity
Γr=Γz=0 (kHz)Γr=40, Γz=0Γr=0, Γz=40Γr=Γz=40
0 4 8 12 16 20Preparation time τ (µs)
1
1.5
2
2.5
3
Mea
n ex
cita
tion
num
ber
<n>
4 8 12 16 200
0.5
1
1.5
Man
del Q
τ
τ
w
Relaxations destroy adiabatic following of the ground state
|R0〉 → |R1〉 → |Rmin2 〉 → |Rmin
3 〉Heraklion, 25/10/17 – p. 22/29
FORTHIESL
FORTHIESL
Simulations of adiabatic dynamics
F = Pmin2 F = Pmin
4
4 8 12 16 20τ
−1−0.9−0.8
Q
1
1.5
2
<n
>
00.10.20.30.40.50.6
F
4 8 12 16 20τ
−1−0.9−0.83
3.5
4
0
0.1
0.2
0.3N=31N=13
Petrosyan, Mølmer, Fleischhauer, JPB 49, 084003 (2016)
Heraklion, 25/10/17 – p. 23/29
FORTHIESL
FORTHIESL
Extended 2D systems
Heraklion, 25/10/17 – p. 24/29
FORTHIESL
FORTHIESL
Rydberg Quasi-Crystals (2D)
alat = 0.532 µm, db = 3.81 µm
Schauß et al., Nature 491, 87 (2012)
Heraklion, 25/10/17 – p. 25/29
FORTHIESL
FORTHIESL
Spatial and Angular Correlations
g(2)(r) =∑
i 6=j δr,rij 〈σirrσ
jrr〉
∑i 6=j
δr,rij 〈σirr〉〈σ
jrr〉
g(2)(∆φ) =∫
dφ2π
∑i,j δφ,φi
δφ+∆φ,φj〈σi
rrσjrr〉
∑i,j
δφ,φi〈σi
rr〉δφ+∆φ,φj〈σj
rr〉
Schauß et al., Nature 491, 87 (2012)
Heraklion, 25/10/17 – p. 26/29
FORTHIESL
FORTHIESL
Rydberg Quasi-Crystals
0
0.05
0 1 2 3 4 5n
0
0.2
0.4
0.6
0.8
1p R
(n)
⟨nR⟩=1.17
0
0.1
0.2
0.3
0.4
CoM
ρ
φx
y
−1 −0.5 0 0.5 1φ/π
00.20.40.60.8
1
P(φ
)
0 1ρ/db
00.20.40.60.8
1
P(ρ
)
bd n=1 n=2
σrr
d=1db
Heraklion, 25/10/17 – p. 27/29
FORTHIESL
FORTHIESL
Rydberg Quasi-Crystals
0 1 2 3 4 5 6 7 8 9 10n
0
0.2
0.4
0.6
0.8
p R(n
)
⟨nR⟩=3.63
0
0.05
−1 −0.5 0 0.5 1φ/π
0
0.2
0.4
0.6
P(φ
)
0 1ρ/db
0
0.2
0.4
0.6
P(ρ
)
−1 −0.5 0 0.5 1φ/π
0 1ρ/db
0
0.05
0.1
0.15
d=2db
bd n=3 n=4
σrr
Heraklion, 25/10/17 – p. 27/29
FORTHIESL
FORTHIESL
Rydberg Quasi-Crystals
0
0.05
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15n
0
0.2
0.4
0.6
p R(n
)
⟨nR⟩=6.62
d=3db
d
−1 −0.5 0 0.5 1φ/π
0
0.2
0.4
P(φ
)0 1 2
ρ/db
0
0.2
0.4
0.6
0.8
1
P(ρ
)
−1 −0.5 0 0.5 1φ/π
0 1 2ρ/db
b n=6 n=7
σrr
Heraklion, 25/10/17 – p. 27/29
FORTHIESL
FORTHIESL
Rydberg Quasi-Crystals
0
0.05
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20n
0
0.2
0.4
p R(n
) ⟨nR⟩=10.4
db=4d
−1 −0.5 0 0.5 1φ/π
0.1
0.2
0.3
P(φ
)
0 1 2ρ/db
0
0.2
0.4
0.6
0.8
1
P(ρ
)
−1 −0.5 0 0.5 1φ/π
0 1 2ρ/db
bd n=10 =11n
σrr
Heraklion, 25/10/17 – p. 27/29
FORTHIESL
FORTHIESL
Rydberg Quasi-Crystals
0
0.05
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20n
0
0.2
0.4
p R(n
) ⟨nR⟩=10.4
db=4d
−1 −0.5 0 0.5 1φ/π
0.1
0.2
0.3
P(φ
)
0 1 2ρ/db
0
0.2
0.4
0.6
0.8
1
P(ρ
)
−1 −0.5 0 0.5 1φ/π
0 1 2ρ/db
bd n=10 =11n
σrr
Petrosyan, PRA 88, 043431 (2013)0 1 2 3 4z/db
0
0.5
1
1.5
g(2) (z
)
zmax
nearly incompressible
Liquid of Rydberg exitations
Heraklion, 25/10/17 – p. 27/29
FORTHIESL
FORTHIESL
Summary: Rydberg states offer
Strong, long-range, switchable interactions between atoms
Heraklion, 25/10/17 – p. 28/29
FORTHIESL
FORTHIESL
Summary: Rydberg states offer
Strong, long-range, switchable interactions between atoms
Logic gates for digital quantum computations and simulations
Heraklion, 25/10/17 – p. 28/29
FORTHIESL
FORTHIESL
Summary: Rydberg states offer
Strong, long-range, switchable interactions between atoms
Logic gates for digital quantum computations and simulations
Analog simulations of few- and many-body quantum systems
Heraklion, 25/10/17 – p. 28/29
FORTHIESL
FORTHIESL
Summary: Rydberg states offer
Strong, long-range, switchable interactions between atoms
Logic gates for digital quantum computations and simulations
Analog simulations of few- and many-body quantum systems
Coupling with other systems for hybrid quantum technologies
PNAS 112, 3866 (2015)
Heraklion, 25/10/17 – p. 28/29
FORTHIESL
FORTHIESL
Thank you!
Heraklion, 25/10/17 – p. 29/29