24.12.2014 Views

Quantum Information Science with Trapped Ca+ Ions with Trapped ...

Quantum Information Science with Trapped Ca+ Ions with Trapped ...

Quantum Information Science with Trapped Ca+ Ions with Trapped ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

<strong>Quantum</strong> <strong>Information</strong> <strong>Science</strong><br />

<strong>with</strong> <strong>Trapped</strong> Ca + <strong>Ions</strong>


<strong>Quantum</strong> <strong>Information</strong> <strong>Science</strong><br />

<strong>with</strong> <strong>Trapped</strong> Ca + <strong>Ions</strong><br />

Rainer Blatt<br />

Institute of Experimental Physics, University of Innsbruck,<br />

Institute of <strong>Quantum</strong> Optics and <strong>Quantum</strong> <strong>Information</strong>,<br />

Austrian Academy of <strong>Science</strong>s<br />

• <strong>Trapped</strong> Ca + for quantum information processing<br />

• Tffli Toffoli gate <strong>with</strong> trapped ions<br />

• Mølmer-Sørensen high fidelity gate operation<br />

• <strong>Quantum</strong> computation <strong>with</strong> logical qubits<br />

• Testing quantum mechanics <strong>with</strong> trapped ions<br />

FWF<br />

CONQUEST<br />

Industrie<br />

IQI<br />

€<br />

SFB SCALA Tirol<br />

GmbH<br />

$


The requirements for quantum computation<br />

D. P. DiVincenzo, Quant. Inf. Comp. 1 (Special), 1 (2001)<br />

I. Scalable physical system, well characterized qubits<br />

II.<br />

III.<br />

Ability to initialize the state of the qubits<br />

Long relevant coherence times, much longer than gate operation time<br />

IV.<br />

“Universal” set of quantum gates<br />

V. Measurement capability specific to implementation<br />

VI.<br />

VII.<br />

Ability to interconvert stationary and flying qubits<br />

Ability to faithfully transmit flying qubits between specified locations<br />

The seven commandments for QC !!


<strong>Quantum</strong> information processing <strong>with</strong> trapped ions<br />

control bit<br />

target bit<br />

other gate proposals (and more):<br />

• Cirac & Zoller<br />

• Mølmer & Sørensen, Milburn<br />

• Jonathan & Plenio & Knight<br />

• Geometric phases<br />

• Leibfried & Wineland


Qubits <strong>with</strong> trapped ions<br />

Storing and keeping quantum information requires long-lived atomic states:<br />

• optical transition frequencies<br />

(forbidden transitions,<br />

intercombination lines)<br />

S – D transitions in alkaline earths:<br />

Ca + , Sr + , Ba + , Ra + , (Yb + , Hg + ) etc.<br />

• microwave transitions<br />

(hyperfine transitions,<br />

Zeeman transitions)<br />

alkaline earths:<br />

9<br />

Be + , 25 Mg + , 43 Ca + , 87 Sr + ,<br />

137<br />

Ba + , 111 Cd + , 171 Yb +<br />

P 1/2<br />

D 5/2<br />

S 1/2<br />

P 3/2<br />

TLS<br />

S 1/2<br />

TLS<br />

S 1/2 Boulder 9 Be + ; Michigan 111 Cd + ;<br />

Innsbruck 40 Ca + Innsbruck 43 Ca + , Oxford 43 Ca + ;<br />

Maryland 171 Yb + ;


Level scheme of Ca +<br />

P 3/2<br />

854 nm<br />

qubit on narrow S - D<br />

quadrupole transition<br />

1<br />

s<br />

P 1/2<br />

393 nm<br />

397 nm<br />

866 nm<br />

D 5/2<br />

D 3/2<br />

729 nm<br />

40 Ca +<br />

S 1/2


Innsbruck linear ion trap (2000)<br />

1.0 mm<br />

6 mm


Detection of 6 individual ions


Spectroscopy of the S 1/2 –D 5/2 transition<br />

P 1/2 D 5/2<br />

2 level system:<br />

Fluorescence<br />

detection<br />

2-level-system:<br />

S 1/2<br />

Zeeman structure in<br />

non-zero magnetic field:<br />

D 5/2<br />

-3/2 -1/2 1/2 3/2<br />

-5/2<br />

3/2<br />

5/2<br />

<br />

sideband<br />

+ vibrational<br />

quantum<br />

cooling<br />

degrees of<br />

state<br />

S 1/2<br />

1/2 freedom processing<br />

-1/2


Quantized Ion Motion<br />

2-level-atom<br />

harmonic trap<br />

coupled system<br />

excitation: various resonances<br />

spectroscopy: carrier and sidebands<br />

n = 0<br />

D 5/2<br />

n = 1<br />

n = -1<br />

S 1/2<br />

n = 0 1 2<br />

<br />

Laser detuning


Coherent state manipulation<br />

D,0<br />

D ,1<br />

carrier<br />

(C)<br />

S,1<br />

S,0<br />

carrier and sideband<br />

Rabi oscillations<br />

<strong>with</strong> Rabi frequencies<br />

blue sideband<br />

(BSB)<br />

Lamb-Dicke parameter


<strong>Quantum</strong> information processing <strong>with</strong> trapped ions<br />

►algorithms:<br />

sequence of single qubit and<br />

two-qubit gate operations<br />

►gate operations:<br />

sequences of laser pulses<br />

(carrier and/or sideband pulses)<br />

carrier sideband<br />

d<br />

►analysis: l i<br />

measure density matrix of<br />

state or process (tomography)<br />

measure entanglement<br />

g<br />

via parity oscillations


Scalable push-button generation of GHZ states<br />

Fidelity<br />

Fidelity


Eight – Ion W state<br />

Fidelity: 0.76<br />

6561 settings,<br />

~ 10 h measurement time,<br />

but reconstruction time:<br />

~ several days on a<br />

computer cluster<br />

genuine<br />

e<br />

8-particle<br />

entanglement !<br />

The<br />

<strong>Quantum</strong> Byte<br />

H. Häffner et al., Nature 438, 643 (2005)


<strong>Quantum</strong> Process Tomography<br />

control bit<br />

target bit<br />

target bit<br />

control bit<br />

characterizes gate operation completely


– matrix for observed CNOT gate operation


– matrix for observed CNOT gate operation


Scalable quantum computation requires error correction


Toffoli gate: controlled-controlled NOT<br />

Toffoli gate (Tommaso Toffoli, 1980):<br />

…… is a universal reversible logic gate, i.e. any reversible circuit can<br />

be constructed from Toffoli gates.<br />

also known as the controlled-controlled-NOT or CCNOT-gate operation<br />

useful, e.g. for error correction


Toffoli gate: pulse sequence<br />

use 2-phonon excitation<br />

Th. Monz et al.,<br />

Phys. Rev. Lett. 102, 040501 (2009)<br />

B B B B B B B B<br />

B<br />

B<br />

effective <br />

for n=1,2<br />

result:<br />

n=1 iff SS<br />

C B B B C<br />

CNOT<br />

iff n=1<br />

undo encoding,<br />

reverse phases


Toffoli gate: experimental truth table<br />

density matrix<br />

State Fidelity<br />

Th. Monz et al., Phys. Rev. Lett. 102, 040501 (2009)


Toffoli gate: process tomography<br />

- matrix for ideal TOFFOLI gate operation<br />

III<br />

IIXIIY<br />

IIYIIZIXI<br />

ZZZ<br />

ZZZ<br />

III


Toffoli gate: process tomography<br />

- matrix for the real TOFFOLI gate operation<br />

Mean<br />

Process Fidelity<br />

Th. Monz et al., Phys. Rev. Lett. 102, 040501 (2009)


<strong>Quantum</strong> procedures and fidelities<br />

► single qubit operations > 99 %<br />

► 2-qubit CNOT gate ~ 93 %<br />

► Bell states<br />

► W and GHZ states<br />

93-95 %<br />

85-90 %<br />

► <strong>Quantum</strong> teleportation 83 %<br />

► Entanglement swapping<br />

~ 80 %<br />

► Toffoli gate operation 71 %<br />

!<br />

BUT: for fault-tolerant operation needed > 99 %


Mølmer - Sørensen gate - operation<br />

bichromatic laser excitation<br />

close to upper and lower<br />

sidebands induces collective<br />

state changes (spin flips)<br />

K. Mølmer, A. Sørensen,<br />

Phys. Rev. Lett. 82, 1971 (1999)<br />

C. A. Sackett et al.,<br />

Nature 404, 256 (2000)


measure the parity :<br />

Measuring entanglement<br />

C. A. Sackett et al., Nature 404, 256 (2000)<br />

Ion 1<br />

Ion 2<br />

/2<br />

<br />

CNOT /2 <br />

/2 <br />

oscillates t <strong>with</strong> 2<br />

54% visibility<br />

Fidelity =<br />

0.5(P SS +P DD +visibility) =<br />

71(3) %<br />

F. Schmidt-Kaler et al., Nature 422, 408 (2003)


Deterministic Bell states using the Mølmer-Sørensen gate<br />

entangled<br />

J. Benhelm, G. Kirchmair,<br />

C. Roos<br />

Theory: C. Roos,<br />

New Journ. of Physics 10,<br />

013002 (2008)<br />

measure entanglement<br />

via parity oscillations<br />

gate duration<br />

average fidelity


Mølmer-Sørensen gate: thermal states<br />

Gate operation after ground Doppler state cooling cooling<br />

Fidelity :<br />

F=98 99.3(1) 98.0(1) %<br />

Gate operation ≈ independent of motional state !<br />

G. Kirchmair et al., New. J. Phys. 11, 023002 (2009)


Collective entangling gates + individual light shifts<br />

N ions<br />

Basic set of operations:<br />

Mølmer-Sørensen gate<br />

individual light shift gates<br />

collective spin flips<br />

+ + + + +<br />

● favorable ion addressing by light shifts (~ 2 )<br />

● no interferometric stability between beams required<br />

Arbitrary unitary operations can be achieved !<br />

...but how


Optimal control for arbitrary quantum gates<br />

<strong>Quantum</strong> optimal control:<br />

V. Nebendahl et al.,<br />

Phys. Rev. A 79,<br />

012312 (2009)<br />

Find<br />

such that<br />

Gradient ascent algorithm: N. Khaneja et al., , J. Magn. Res. 172, , 296 (2005)<br />

Modification of search algorithm:<br />

Example: quantum Toffoli gate<br />

● no simultaneous application of several Hamiltonians<br />

● sequence of pulses <strong>with</strong> variable length<br />

1<br />

2<br />

=<br />

3


Optimal control : <strong>Quantum</strong> Error Correction<br />

<strong>Quantum</strong> Error Correction: 3 qubits encode logical qubit (protection against spin flips)<br />

spin flip errors<br />

spin flip errors<br />

encoding<br />

reset ancillas<br />

error syndrome<br />

detection<br />

correction<br />

step<br />

Implementation : 34 laser pulses (11 entangling pulses)<br />

V. Nebendahl et al.,<br />

Phys. Rev. A 79,<br />

012312 (2009)


Scalability is facilitated by error avoidance<br />

Decoherence-free Bell states:<br />

C. Roos et al., Phys. Rev. Lett. 92, 220402 (2004)<br />

decoherence-time:<br />

0.5 x 1.05(15) s


prepare qubits<br />

in S states<br />

t<br />

D 1/2 3/2<br />

-1/2<br />

Robust entanglement<br />

5/2<br />

H. Häffner et al., Appl. Phys. B 81, 151 (2005)<br />

Hiding states in S, S‘ states<br />

avoids decoherence from<br />

-5/2 -3/2 spontaneous emission<br />

D 5/2<br />

S 1/2<br />

S<br />

-1/2<br />

1/2<br />

S‘


<strong>Quantum</strong> gates <strong>with</strong>in a decoherence free subspace<br />

Idea: use long-lived Bell states as logical qubits<br />

K. Kim et al., Innsbruck, 2006<br />

Gate operations by<br />

- addressing individual ions<br />

- simultaneous addressing of innermost ions<br />

Trade off more ions (2x) for much extended coherence times (100-1000x)


Universal quantum computation <strong>with</strong> logical qubits<br />

Idea: use long-lived Bell states as logical qubits<br />

Th. Monz, K. Kim,<br />

Ph. Schindler (2008)<br />

Gate operations by<br />

- addressing individual ions and MS-gates<br />

- simultaneous addressing of innermost ions<br />

1 st experiment:<br />

Analysis by state and process tomography


Gate operations <strong>with</strong> logical qubits<br />

Z<br />

Z<br />

K. Kim et. al.<br />

Phys. Rev. A77, 050303 (2008)<br />

X<br />

C. Roos,<br />

New Journ. Phys. 10 (2008)<br />

Z<br />

single qubit phase shift by<br />

AC Stark shifts<br />

F. Schmidt-Kaler et. al.<br />

Europhys. Lett. 65, 587 (2004)


CNOT gate operation <strong>with</strong> a logical qubit<br />

spin echo


theory<br />

Process tomography of CNOT <strong>with</strong> logical qubits


Process tomography of CNOT <strong>with</strong> logical qubits<br />

experiment<br />

mean gate fidelity


<strong>Quantum</strong> state manipulation <strong>with</strong> logical qubits<br />

results and limitations:<br />

● mean gate fidelity: 89(4)%<br />

● limited by addressing errors, spurious laser noise, position stability,<br />

equal illumination of neighbouring ions<br />

advantages:<br />

● insensitive to laser linewidth,<br />

● insensitive to AC-Stark shifts<br />

(work in progress)<br />

● lifetime limited coherence time<br />

(work in progress)


Testing quantum mechanics <strong>with</strong> trapped ions<br />

• Do observables have predetermined values,<br />

fixed by hidden variables<br />

J. S. Bell<br />

“No physical theory<br />

of local hidden<br />

variables can<br />

reproduce all of<br />

the predictions of<br />

quantum mechanics” S. Kochen E. Specker<br />

“Non‐contextual t lhidden<br />

variable theory cannot<br />

reproduce the predictions of<br />

quantum mechanics”


Kochen‐specker theorem<br />

“Non‐contextual hidden variable theory cannot<br />

reproduce the predictions of quantum mechanics”<br />

Non‐Contextuality<br />

Any measurement A should give a value independent<br />

of other compatible measurements B :<br />

(AB) = (A) ∙ (B)<br />

Experimental tests:<br />

Huang, Y.F., Phys. Rev. Lett. 90, 250401 (2003)<br />

Hasegawa, Y., Phys. Rev. Lett. 97, 230401 (2006)<br />

Specker, R., Dialectica 14, 239‐246 (1960)<br />

Bell, J.S., Rev. Mod. Phys. 38, 447‐452 (1966)<br />

Kochen, S., & Specker, E.P., J. Math. Mech. (1967)


The Peres – Mermin square<br />

Measure observables<br />

A ij<br />

<strong>with</strong> v(A ij<br />

) = ±1,<br />

where the observables<br />

in each row and column<br />

are mutually compatible<br />

A. Peres, Phys. Lett. A 151, 107–108 (1990)<br />

N. D. Mermin, Phys. Rev. Lett. 65, 3373–3376 (1990)<br />

<br />

(1)<br />

z <br />

(2)<br />

z <br />

(1)<br />

z <br />

(2)<br />

z R 1 =1<br />

1<br />

z z z z<br />

(2) x <br />

(1)<br />

x <br />

(1)<br />

x <br />

(2)<br />

x R 2 =11<br />

z<br />

(1) x<br />

(2)<br />

x<br />

(1) z<br />

(2)<br />

y<br />

(1) y<br />

(2)<br />

R 3 =1<br />

C 1 =1 C 2 = 1 C 3 = ‐1


Testing the Kochen‐Specker theorem<br />

● KS = R 1 + R 2 + R 3 + C 1 + C 2 ‐ C 3 <br />

● Non‐contextual hidden variable theory<br />

KS ≤ 4<br />

● quantum mechanics<br />

KS =6<br />

<br />

(1) z <br />

(2) z <br />

(1) <br />

(2)<br />

z z<br />

R 1 =1<br />

for any input state <br />

(2)<br />

x <br />

(1)<br />

x <br />

(1)<br />

x <br />

(2)<br />

x<br />

R 2 =1<br />

z<br />

(1) x<br />

2<br />

x<br />

(1) z<br />

(2)<br />

y<br />

(1) y<br />

(2)<br />

R 3 =1<br />

C 1 =1 C 2 = 1 C 3 = ‐1<br />

A. Cabello, Phys. Rev. Lett. 101, 210401 (2008)


Measuring operators like x x<br />

(1) (2) (1) (2)<br />

z z z z<br />

(2) x <br />

(1)<br />

x <br />

(1)<br />

x <br />

(2)<br />

x<br />

z<br />

(1) x<br />

(2)<br />

x<br />

(1) z<br />

(2)<br />

y<br />

(1) y<br />

(2)<br />

We have to measure all observables <strong>with</strong> the same setup<br />

whether measured in a row or a column map to single ion


Experimental procedure<br />

Measurement scheme for rows and columns of PM square:<br />

unitary for respective<br />

element of PM square<br />

actual measurement


Results<br />

z<br />

(1)<br />

z<br />

(2)<br />

z<br />

(1) z<br />

(2) R 1 =0.92(1)<br />

Input state:<br />

|SS |DD<br />

<br />

(2)<br />

x <br />

(1)<br />

x <br />

(1)<br />

x <br />

(2)<br />

x R 2 =0.93(1) 0 z<br />

(1) x<br />

2<br />

x<br />

(1) z<br />

(2)<br />

y<br />

(1) y<br />

(2) R 3 =0.90(1)<br />

C =0.90(1) C = 0.89(1)C = ‐0.91(1) 1 2 3 = 5.46(4)<br />

KS > 4


State independence<br />

States are prepared <strong>with</strong> 97(2) % fidelity (quantum state tomography)<br />

KS avg = 5.38(4)<br />

> 4<br />

State independence


Simulating the Dirac equation<br />

… <strong>with</strong> a single trapped ion


Simulating the Dirac equation<br />

Lamata et al., PRL 98, 253005 (2007)<br />

The Dirac equation<br />

can be cast in a 1+1 dimensional form (1 spatial + 1 spin degree of freedom)<br />

resonant bichromatic excitation<br />

Stark shift<br />

<strong>with</strong> the replacements


<strong>Quantum</strong> <strong>Information</strong> <strong>Science</strong> <strong>with</strong> <strong>Trapped</strong> <strong>Ions</strong><br />

Ca + for quantum information processing<br />

<strong>Quantum</strong> process tomography, gate operations<br />

Toffoli gate <strong>with</strong> trapped ions<br />

Mølmer-Sørensen high fidelity gate operation<br />

<strong>Quantum</strong> gate operations <strong>with</strong> logical qubits<br />

Testing quantum mechanics <strong>with</strong> trapped ions<br />

Simulating the Dirac equation <strong>with</strong> a trapped ion<br />

Future:<br />

further optimization of logic operations<br />

error correction protocols <strong>with</strong> three and five qubits<br />

implementation <strong>with</strong> 43 Ca + , logical qubits + scalability<br />

miniaturize traps, interface quantum information


Future goals and developments<br />

more qubits (~20 – 50)<br />

better fidelities<br />

faster gate operations<br />

faster detection<br />

cryogenic trap, micro-structured traps<br />

development of 2-d trap arrays, onboard addressing, electronics etc.<br />

entangling of large(r) systems: characterization <br />

implementation of error correction, keep „qubit alive“<br />

applications<br />

- small scale QIP (e.g. repeaters)<br />

- quantum metrology, enhanced S/N, tailored atoms and states<br />

- quantum simulations<br />

- quantum computation


The international Team 2008<br />

€<br />

FWF<br />

SFB<br />

CONQUEST<br />

SCALA<br />

Industrie<br />

Tirol<br />

IQI<br />

GmbH<br />

$


The international Team 2008<br />

C. Roos<br />

H. Häffner<br />

W. Hänsel<br />

P. Schmidt<br />

M. Hennrich<br />

R. Gerritsma<br />

F. Dubin<br />

K. Kim<br />

A. Villar<br />

M. Brownnutt<br />

N. Daniidilis<br />

T. Northup<br />

J. Barreiro<br />

G. Hetet<br />

M. Riebe<br />

J. Benhelm<br />

C. Russo<br />

H. Barros<br />

F. Splatt<br />

S. Narayanan<br />

F. Zähringer<br />

P. Schindler<br />

B. Brandstädter<br />

M. Kumph<br />

A. Stute<br />

M. Harlander<br />

B. Hemmerling<br />

M. Chwalla<br />

Th. Monz<br />

G. Kirchmair<br />

S. Gerber<br />

L. Slodička<br />

M. Niedermair<br />

L. Anderlan<br />

V. Nebendahl<br />

R. Lechner<br />

D. Nigg<br />

€<br />

FWF<br />

SFB<br />

CONQUEST<br />

SCALA<br />

Industrie<br />

Tirol<br />

IQI<br />

GmbH<br />

$

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!