Page 1
Hartmut HäffnerInstitute for Quantum Optics and Quantum InformationInnsbruck, Austria
Quantum computing with trapped ions
Berkeley, Nov 25th 2008
€$
SCALAQGATES
FWF SFB
IndustrieTirol
IQIGmbH
• Basics of ion trap quantum computing
• Measuring a density matrix
• Quantum gates
• Deutsch Algorithm
Page 2
000001010011
100011110011
Quantumprocessor
Cavity QED
Superconducting qubits
Trapped ions
Quantum dots
NMR
© A. Ekert
Which technology ?
Page 3
000001010011
100011110011
Quantumprocessor
Cavity QED
Superconducting qubits
Trapped ions
Quantum dots
NMR
© A. Ekert
Which technology ?
Page 4
Good things about ion traps:
Ions are excellent quantum memories; single qubit coherence times > 10 minutes have been demonstrated (Boulder 1991)
(
Ions can be controlled very well
Many ideas to scale ion traps
Bad things about ion traps:
Slow (~1 MHz)
Technically demanding
Why trapped ions ?
Page 5
The hardware
S1/2
P1/2
D5/2
Qubit
Page 6
Trapped ions form the quantum register
Innsbruck quantum processor
Trap electrodes
Page 7
• Scalable physical system, well characterized qubits
• Ability to initialize the state of the qubits
• Long relevant coherence times, much longer than gate operation time
• “Universal” set of quantum gates
• Qubitspecific measurement capability
DiVincenzo criteria
Page 8
1. Initialization in a pure quantum state P1/2 D5/2
=1s
S1/2
40Ca+
P1/2
S1/2
D5/2P1/2
S1/2
D5/2D5/2
S1/2
Experimental procedure
Page 9
1. Initialization in a pure quantum state P1/2 D5/2
=1s
S1/2
40Ca+
P1/2
S1/2
D5/2
Dopplercooling Sideband
cooling
P1/2
S1/2
D5/2D5/2
S1/2
Experimental procedure
Page 10
2. Quantum state manipulation on S1/2 – D5/2 transition
Quantum statemanipulation
1. Initialization in a pure quantum state P1/2
S1/2
D5/2D5/2
S1/2
Experimental procedure
Page 11
P1/2 D5/2
=1s
S1/2
40Ca+
1. Initialization in a pure quantum state:
2. Quantum state manipulation on S1/2 – D5/2 transition
P1/2
S1/2
D5/2
Dopplercooling Sideband
cooling
P1/2
S1/2
D5/2
Quantum statemanipulation
P1/2
Fluorescencedetection
3. Quantum state measurement by fluorescence detection
S1/2
D5/2
Experimental procedure
Page 12
P1/2 D5/2
=1s
S1/2
40Ca+
1. Initialization in a pure quantum state:
2. Quantum state manipulation on S1/2 – D5/2 transition
P1/2
S1/2
D5/2
Dopplercooling Sideband
cooling
P1/2
S1/2
D5/2
Quantum statemanipulation
P1/2
Fluorescencedetection
3. Quantum state measurement by fluorescence detection
S1/2
D5/2
Experimental procedure
5µm
50 experiments / s
Repeat experiments100200 times
Spatially resolveddetection withCCD camera
Two ions:
Page 13
Ds
tate
pop
ulat
ion
Rabi oscillations
P1/2
S1/2
D5/2D5/2
Page 14
Ds
tate
pop
ulat
ion
Rabi oscillations
P1/2
S1/2
D5/2D5/2
Page 15
Ds
tate
pop
ulat
ion
Rabi oscillations
P1/2
S1/2
D5/2D5/2
Page 16
The phase ...
+
++ =
+
++e =ωt
Page 19
CCD
Paul trap
Fluorescencedetection
electrooptic deflector
coherentmanipulation of qubits
dichroicbeamsplitter
inter ion distance: ~ 4 µm
addressing waist: ~ 2 µm
< 0.1% intensity on neighbouring ions
10 8 6 4 2 0 2 4 6 8 100
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Exc
itatio
nDeflector Voltage (V)
D
Addressing single qubits
Page 20
Memory errors:
Bitflips
Dephasing
Operationial errors
technical imperfections …
Decoherence mechanisms
Page 21
Ramsey Experiment
= 15.9 ms
π / 2 π / 2Ramsey Time
Dephasing of qubits
Page 23
106 s
101 s
103 s
104 s
103 s
101 s
Single qubit gates
Two qubit gates (Geometric phase gates)
T
Single qubit coherence (magnetic field sensitive)
S
Coherence of the motion
105 s
102 s
100 s
102 s
Two qubit gates (CiracZoller approach)
T
Realized time scales
Page 24
qubit
Raman transitions:
Excited state
Ground state
Long lived qubits
Page 25
Raman transitions:
Excited state
Ground state
Long lived qubits
Page 26
From: C. Langer et al., PRL 95, 060502 (2005), NIST
Level scheme of 9Be+:
Long lived qubits
Page 27
From: C. Langer et al., PRL 95, 060502 (2005), NIST
Long lived qubits
Page 28
106 s
101 s
103 s
104 s
103 s
101 s
Single qubit gates
Two qubit gates (Geometric phase gates)
T
Single qubit coherence (magnetic field sensitive)
S
Coherence of the motion
105 s
102 s
100 s
102 s
Two qubit gates (CiracZoller approach)
T
Realized time scales
Single qubit coherence (magnetic field insensitive)
S
Single qubit coherence (magnetic field insensitive + RF drive)
S
Page 29
The common motionacts as the quantumbus.
Having the qubits interact
Page 30
50 µm
The common motionacts as the quantumbus.
Having the qubits interact
Page 31
harmonic trap
……
Ion motion
Page 32
harmonic trap
…
2levelatom joint energy levels
…
Ion motion
Page 33
carrier
carrier and sidebandRabi oscillationswith Rabi frequencies
LambDicke parameter
Coherent manipulation
Page 34
• Introduction to quantum information
• Entangled states
• Teleportation
• Scaling of ion trap quantum computers
• Wiring up trapped ions
Page 35
…
… …
…
Generation of Bell states
Page 36
π
…
… …
…
Generation of Bell states
Page 37
π/2
…
… …
…
Generation of Bell states
Page 38
π…
… …
…
Generation of Bell states
Page 39
π…
… …
…
Bell states with atoms
9Be+: NIST (fidelity: 97 %)
40Ca+: Oxford (83%)
111Cd+: Ann Arbor (79%)
25Mg+: Munich
40Ca+: Innsbruck (99%)
Generation of Bell states
Page 40
Fluorescencedetection withCCD camera:
Coherent superposition or incoherent mixture ?
What is the relative phase of the superposition ?
Analysis of Bell states
Measurement of the density matrix:SSSDDS
DD SSSDDSDD
Page 41
A measurement yields the zcomponent of the Bloch vector
=> Diagonal of the density matrix
Measuring a density matrix
Page 42
A measurement yields the zcomponent of the Bloch vector
=> Diagonal of the density matrix
Rotation around the x or the yaxis prior tothe measurement yields the phase informationof the qubit.
Measuring a density matrix
Page 43
A measurement yields the zcomponent of the Bloch vector
=> Diagonal of the density matrix
=> coherences of the density matrix
Rotation around the x or the yaxis prior tothe measurement yields the phase informationof the qubit.
Measuring a density matrix
Page 44
SSSD
DSDD SSSDDSDD
SSSD
DSDD SSSDDSDD
SSSD
DSDD SSSDDSDD
Decoherence properties of qubits
Page 45
Bellstate survives
Roos et al., Science 304, 1478 (2004)
Measurementof the center ion
A “real” thought experiment
Page 46
Generalized Bell states
Page 47
656100 measurements~ 10 h measurement time
Genuine 8particle entanglement
Häffner et al., Nature 438, 643 (2005)
Generalized Bell states
Page 49
control target
...allows the realization of a universal quantum computer !
Having the qubits interact
Page 50
control target
...allows the realization of a universal quantum computer !
Having the qubits interact
Most popular gates: CiracZoller gate (SchmidtKaler et al., Nature 422, 408 (2003)). Geometric phase gate (Leibfried et al., Nature 422, 412 (2003)). MølmerSørensen gate (Sackett et al., Nature 404, 256 (2000)).
Page 51
Control bit
Target bit
Target
A controlledNOT operation
Page 52
A controlledNOT operation
Ion 1:
Vibration:
Ion 2:
SWAP1SWAPControl qubit
Target qubit
Page 53
A controlledNOT operation
Ion 1
Vibration
Ion 2
SWAP1SWAPControl qubit
Target qubit
Ion 1
Ion 2
Pulse sequence:Laser frequencyPulse lengthOptical phase
Page 54
1
2
3
4
Composite phase gate (2π rotation)
Page 55
4
3
2
1
Action on |S,1> |D,2>
Page 56
0=ϕ 0=ϕ 2πϕ =
0 20 40 60 80 100 120 140 160 1800
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Time (µs)
D5/
2 e
xcita
tion
Single ion composite phase gate
Page 57
Time (µs)
T
D5/
2 e
xcita
tion
0 50 100 150 200 250 3000
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Phase gate
0.978 (5)
π/2π/2
Single ion composite CNOT gate
Page 58
A controlledNOT operation
Ion 1
Vibration
Ion 2
SWAP1SWAPControl qubit
Target qubit
Ion 1
Ion 2
Pulse sequence:Laser frequencyPulse lengthOptical phase
Page 59
OutputInput
Probability
Truth table of the CNOT
Page 60
prepare CNOT detect
Using a CNOT to create a Bell state
output
Page 61
Raman transitions between
Interaction of two ions via common motion.
n=0
n=1
MølmerSørensen gate
Page 62
Raman transitions between
Interaction of two ions via common motion.
MølmerSørensen gate
Page 63
Raman transitions between
Interaction of two ions via common motion.
MølmerSørensen gate
Page 64
bicromatic beam applied to both ions
Technical realization
Page 65
entangled ?
Entangling ions
J. Benhelm et al., Nature Physics 4, 463 (2008)
Theory: C. Roos, NJP 10, 013002 (2008)
Page 66
gate duration
average fidelity: 99.3 (2) %
entangled
measure entanglementvia parity oscillations
Entangling ions
J. Benhelm et al., Nature Physics 4, 463 (2008)
Theory: C. Roos, NJP 10, 013002 (2008)
Page 68
maximally entangled states
Gate concatenation
Page 70
The MS gate operation is optimized by
• pulse shaping, minimizes offresonant excitation
• phase relation of amplitude modulated dualfrequency beam
C. Roos, NJP 10, 013002 (2008).
Optimization of the gate
Page 72
• Scalable physical system, well characterized qubits / ?
• Ability to initialize the state of the qubits
• Long relevant coherence times, much longer than gate operation time
• “Universal” set of quantum gates
• Qubitspecific measurement capability
Often neglected:
• exceptional fidelity of operations
• low error rate also for large quantum systems
• all requirements have to met at the same time
DiVincenzo criteria
Page 73
Cirac, Zoller, Kimble, Mabuchi Zoller, Tian, Blatt
Scaling of ion trap quantum computers
Its easy to have thousands of coherent qubits …but hard to control their interaction
Kielpinski, Monroe, Wineland
Page 74
The Michigan T trap
Page 75
An implementation of the Deutschalgoritm …
Page 76
Deutsch‘s problem: Introduction
Decide which class the coin is: False (equal sides) or Fair
A single measurement does NOT give the right answer
Front
Back
Page 77
Deutsch‘s problem: Mathematical formulation
4 possible coins are representend by 4 functions
false
fair0110f(1)
(
10 10f(0)
(
Case 4Case 3Case 2Case 1
BalancedConstant
Page 78
false
fair
ZCNOTCNOTNOTIDz + f(x)
)
0110f(1)
(
10 10f(0)
(
Case 4Case 3Case 2Case 1
BalancedConstant
Deutsch‘s problem: Mathematical formulation
4 possible coins are representend by 4 functions
Physically reversible process realized by a unitary transformation+
Page 79
Deutsch Jozsa quantum circuit
f1
f2
f3
f4
1000010000100001
1000010000010010
0100100000100001
0100100000010010
Ufn
x x
z f(x) + z
NOT
ID
CNOT
ZCNOT
Case Logic Quantum circuit Matrix
Page 80
Quantum analysis gives the right answer after a single measurement!
•D. Deutsch, R. Josza, Proc. R. Soc. London A439, 553 (1992)
D
•M. Nielsen, I. Chuang, QC and QI, Cambridge (2000)
M
Uf
x
z
x
z + f(x)
z
|0>
|1>
Deutsch Jozsa quantum circuit
π/2
π/2
π/2
π/2
Page 81
Uf
x
z
x
z + f(x)
z
|0>
|1>
No information in the second qubit
electronic qubit
motional qubit
π/2
π/2
π/2
π/2
Page 82
D5/2
729 nm
|1>
|0>
internal qubit
Qubits in 40Ca+
S1/2
motional qubit
|0>|1>1
n=0
2
computational subspace
|S,0>
|D,0>|D,1>
|S,1>
Page 83
Uf
x
z
x
z + f(x)
z
|0>
|1>
No information in the second qubit
electronic qubit
motional qubit
π/2
π/2
π/2
π/2
Page 84
Deutsch Jozsa quantum circuit
f1
f2
f3
f4
1000010000100001
1000010000010010
0100100000100001
0100100000010010
Ufn
x x
z f(x) + z
NOT
ID
CNOT
ZCNOT
Case Logic Quantum circuit Matrix
Page 85
Deutsch Jozsa: Realization
x x
z f(x) + z
Page 86
Deutsch Jozsa: Realization
x x
z f(x) + z
Swap Swap
Page 87
|S,0>
|D,0>|D,1>
|S,1>
3step composite SWAP operation
1
2
3
1
3
I. Chuang et al., Innsbruck (2002)
I
Page 88
Deutsch Jozsa: Realization
x x
z f(x) + z
Swap Swap
Page 89
Deutsch Jozsa: Realization
x x
z f(x) + z
Swap Swap
Phasegate
Page 90
1
2
3
4
Composite phase gate (2π rotation)
Page 91
4
3
2
1
Action on |S,1> |D,2>
Page 92
Deutsch Jozsa: Realization
x x
z f(x) + z
Swap Swap
Phasegate
Page 93
Deutsch Jozsa: Realization
x x
z f(x) + z
Swap Swap
Phasegate
Page 94
Deutsch Jozsa: Realization
x x
z f(x) + z
Swap Swap
Phasegate
Phasegate
Page 95
Deutsch Jozsa: Realization
Page 96
Deutsch Jozsa: Realization
Time (µs)
T
Page 97
Deutsch Jozsa: Result
0.986(4)
0
0.931(9)
0
0.90(1)
0
measured |<1|w>|2
1111expected |<1|w>|2
0.975(2)
0
0.975(4)
0
0.087(6)
0
0.019(6)
0
measured |<1|a>|2
1100expected |<1|a>|2
Case 4Case 3Case 2Case 1
BalancedConstant
S. Gulde et al., Nature 412, 48 (2003)
Page 98
• Basics of ion trap quantum computing
• Measuring a density matrix
• Quantum gates
• Deutsch Algorithm
Conclusions
Berkeley, Nov 25th 2008
€$
SCALAQGATES
FWF SFB
IndustrieTirol
IQIGmbH