Quantum information processing in ion traps II
D. J. Wineland, NIST, Boulder
Part I, Rainer Blatt
Lecture 1: Nuts and bolts Ion trapology Qubits based on ground-state hyperfine levels Two-photon stimulated-Raman transitions * Rabi rates, Stark shifts, spontaneous emission Lecture 2: Quantum computation (QC) and quantum-limited measurement Trapped-ion QC and DiVincenzos criteria Gates Scaling Entanglement-enhanced quantum measurement Lecture 3: Decoherence Memory decoherence Decoherence during operations * technical fluctuations * spontaneous emission * scaling Decoherence and the measurement problem
Ion trapping 101
Earnshaws theorem: In a charge free region, cannot confine a charged particle with static electric fields. Proof: For confinement, must have (2(q)/2xi)trap location < 0 (xi {x,y,z}) But from Laplaces equation: 2 = 0, cannot satisfy confinement condition for all xi. B0
Solution 1: Penning trap: q U0 [2z2 x2 y2] Difficult to accomplish individual ion addressing.
(However, see: Ciaramicoli, Marzoli, Tombesi, PRL 91, 017901 (2003))
U0
Solution 2: RF-Paul trap: = (x2 + y2 + z2)V0cost + U0(x2 + y2 + z2) [ + + = + + = 0] (Laplace)
Special case: linear RF trap
(quadrupole mass filter plugged on-axis With static fields)
~2R
V0 cos Tt Uo Uc
y
x
axis trap z
end view
(Get s and numerically)
Equations of motion (classical treatment adequate)
Mathieu equation (2)
z-motion, qz = 0 (static harmonic well)
x,y motion, Mathieu equation:
plug into (2), find (recursion relation for) C2n
Solution in ith direction (i {x,y}):
1.0 .8 .6 .4 .2 ai 0 -.2 -.4 -.6
static potential strength
unstable
stable
unstable
.2
.4
.6
.8
1.0
pseudo-potential strength
|qi|
Simultaneous solution for x, y, z:
ai Ui,
ax + ay + az =0
1.0 .8 .6 .4 .2 ax 0 -.2 -.4 -.6 -.8 -1.0 -1.2 0
x = T/2 y 0 stable
az
x 0 y = T/2
-1.0 -.8 -.6 -.4 -.2 0 ay .2 .4 .6 .8 1.0
.2
.4
.6
.8
1.0
|q |
0.05 typically, we live here |ai| , qi2 << 1
ax
ay
-0.05
0.1
qi
0.2
secular motion
micromotion
Heuristic approach: pseudo-potential approximation
assume mean ion position changes negligibly in duration 2/T
pseudo-potential from micromotion kinetic energy: agrees with Mathieu equation limit |ai| , qi2 << 1
for linear RF trap,
Digression: Optical dipole traps:
atomic core
L 0
Response of atomic core to laser field is negligible because of heavy mass. For L >> 0 (blue detuning) electron response out of phase with electric force. Electron trapped in ponderomotive (pseudo-potential) laser potential (just like RF trap). Trapping in field minima. Core is attached to electron Dispersion: For L << 0 (red detuning) electron response in phase with electric force. Trapping in field maxima.
outer electron
(some) ion-trap realities
Patch potentials: Static potentials: pushes ions away from trap axis micromotion xsinTt can cause X-tal heating Fluctuating patch fields: causes heating; COM primarily affected Source: unknown! (mobile electrons on oxide layers,.. ??)
axis trap z
EDC
Idealized trap:
control electrodes
RF electrodes
Approximation:
3 1 2 4
200 m
IONS
3' 4' 2' 1'
gold-coated alumina wafers
linear Paul (RF) trap VRF ~ 500 V RF ~ 50 250 MHz
Chris Myatt et al.
0.2 mm
~ 1 cm
For 9Be+, V0 = 500 V, T/2 = 200 MHz, R = 200 m x,y/2 ~ 6 MHz
0.2 mm
Ion Trap QC: Proposal: J. I. Cirac and P. Zoller,
PRL 74, 4091 (1995)
Motion data bus
(e.g., center-of-mass mode)
Laser beam
n=3 n=2 n=1 n=0
Internal-state qubit
Stay in two lowest motional states (motion qubit)
n'
ladder states for selected motional mode
n
2 1 0
single photon need good laser frequency stability memory and gate coherence limited by upper state lifetime (~ seconds)
0 = optical frequency:
0 = RF/microwave frequency
memory coherence limited by upper-state lifetime (>> days) sideband transitions weak at RF 2-photon optical stimulated-Raman transitions frequency stability = RF modulator stability vary sideband coupling (Lamb-Dicke parameter) with k2 k1 gate decoherence: spontaneous-Raman scattering (fundamental limit)
E - E =0
Stimulated-Raman transitions: Simple case: Motion: 1-D harmonic well (frequency M), Internal states: 3-level system
electric dipole transitions
r = red
,1 ,0
G kr , r
G kb , b
e,1 e,0
b = blue
e
e >> >> 0 >> M
0 M
,2 ,1 ,0
r = red
,1 ,0
G kr , r
G kb , b
e,1 e,0
b = blue
0 M
,2 ,1 ,0
zero-point wavefunction spread
(+ rotating wave approximation)
similarly:
Adiabatic elimination:
make ansatz:
r = red
,1 ,0
G kr , r
G kb , b
e,1 e,0
b = blue
0 M
,2 ,1 ,0
Stark shift of from blue laser
Add in other Stark shifts
absorb Stark shifts into wave function amplitudes
near a resonance:
Rabi flopping
(Lamb-Dicke parameter) Be+: P = 1 mW, w0 = 25 m, /2 = 100 GHz, /2 ~ 0.5 MHz
Carrier transitions:
Debye-Waller factor Sideband transitions: n = n 1 (n> = larger of n and n) red sideband (n = n-1): can get from Jaynes-Cummings Hamiltonian from cavity-QED
(see, e.g., Raimond, Brune, Haroche, Rev. Mod. Phys. 73, 565 (01))
Complete picture:
Sum over excited states, typically:
2P 3/2 2P 1/2
2S 1/2
can tune out differential Stark shifts can tune out polarization sensitivity For N ions, consider effects of 3N modes Debye-Waller factors from spectator modes
sideband transitions: interference from two-mode transitions: e.g. np -mr = M
Spontaneous emission:
r = red
,1 ,0
G kr , r
e
b = blue
e,1 e,0
G kb , b
0 M
,2 ,1 ,0
Quantum information processing in ion traps II
D. J. Wineland, NIST, Boulder Lecture 1: Nuts and bolts Ion trapology Qubits based on ground-state hyperfine levels Two-photon stimulated-Raman transitions * Rabi rates, Stark shifts, spontaneous emission Lecture 2: Quantum computation (QC) and quantum-limited measurement Trapped-ion QC and DiVincenzos criteria Gates Scaling Entanglement-enhanced quantum measurement Lecture 3: Decoherence Memory decoherence Decoherence during operations * technical fluctuations * spontaneous emission * scaling Decoherence and the measurement problem