Atomic Frequency Standards

NICT Space‐Time Standards GroupNICT Space Time Standards Group

Tetsuya Ido

Clock using real CsClock using real Csx

τ = x / v

v

1 / τ

As longer time atom interacts with RF, the width will be narrower.

Cs beam clockCs beam clock

Optically‐pumped Cesium Beam clock

Cs fountain clock

Cesium Atomic Fountain NICT-CsF1（2）

700

Captured by MOT in (0,0,1) cooling geometry

State-select just above laser cooling regionThree-layers

magneticshields

C-f ield coil

j g g

Rectangular cavity for state-selection

16

20

Microwavecavity

Cylindrical cavity for Ramsey resonance

Detection region above laser cooling region

450

Selectioncavity

Detect ionzone

Detection region above laser cooling region

Three-layers magnetic shieldTrapp ing chamber

Ultra high vacuum of less than 2x10-7Pa

Frequency Stability of NICT-CsF1

4×10‐13 ＠1秒

-3 -2 -1 0 1 2 3

0.9Hz 1.4×10‐15 ＠1日

↓

1日前と1日後では1.4×10‐15

しか値はずれない

10-13

viat

ion

-150 -100 -50 0 50 100 150F 9 192 631 770 (H )

4×10‐13/√τ

冷却温度：1～2μK打上げ速度 4 /

10-14

stan

dard

devFrequency - 9 192 631 770 (Hz)

打上げ速度：4m/sec

打上げ高さ：82cm

ドリフト時間：570ms

↓

10-15

Alla

n s

CsF1-HM

↓

線幅：＜0.9Hz 1 10 100 1000 10000 100000Averaging time τ (sec)

Which oscillator do you like?Which oscillator do you like?(a)

(b) Is drifting, but (a) is not.Then, obviously (a)?

Cs clock

(b) But we often say (a) is noisy and (b) is smooth.

H‐maser

Hydrogen maser

H per fine splitting of h drogenHyper fine splitting of hydrogen

F=0 – F=1 , 1.42GHz

Active clock

Oscillator = standards

All Cs clock are passive. Source oscillator does not have Cs inside.

Components of passive clockp p

2) Precision atomic spectroscopy

ΔDetector

Feedback SystemLocks Oscillator to atomic resonance

) p py

Δνatomic resonance

Local Oscillator νaLocal Oscillator

AtomsOscillator with less phase noise (VCXO, 

1) Hi hl bl ill

VCSO, etc…)

1) Highly stable oscillators

1/9192631770     thdividerdivider

1 pulse per second

Local oscillator for microwave clockLocal oscillator for microwave clock

Cryogenic Sapphire Oscillator

・Sapphire crystal inside liquid heliumSapphire crystal  inside liquid helium

・Whispering Gallery Mode

・Q‐value = 109

・Narrow BPF

NICTのCSONICTのCSO

WGH16,0,0

Q‐value = 1.7×109

11.2005GHz@7K

NICT-CSO(1)

Microwave Copper heat sink

transmission line

Mylar seal

Crimped vacuum seal

Microwave Outer can

Vacuum ~10-6 Torr

cable loop

Inner can

NICT-CSO

1E-12

Nicole vs Anritus#3 (at NICT)Nicole vs Stacy (at UWA)

5071A

1E-13

Nicole vs Stacy (at UWA)

1E-14

SR

AV H‐maser

1E-15

CSO

1 10 100 1000 10000 1000001E-16

Averaging Time (s)

O ti l Cl kOptical Clocks

Accuracy: Optical clock now comparable or even better than Cs1E 9

1E 11

1E-10

1E-9

Essen's Cs clock I2/HeNeH

1E-12

1E-11

Hcerta

inty

Ca

1E-14

1E-13

Yb+

H

H

ratio

nal U

nc Cs redefinitionof the second

Hg+,Yb+

1E-16

1E-15

Hg+

Sr+ Cs μ-wave clock Optical clock Sr Lattice clock

Fr

1950 1960 1970 1980 1990 2000 20101E-17

Year

N H i h b tt th C My talk could be biased to•Now a Hg ion has better accuracy than Cs.•Short time Instability: Cs > ions > lattice clock•Are lattice clocks really promissing?

My talk could be biased to lattice clocks to introduce ideas and latest results

Optical Clock Componentsp p

2) Precision atomic spectroscopy

ΔDetector

Feedback SystemLocks Oscillator to atomic resonance

) p py

Δνatomic resonance

Clock Oscillator

1) Highly stable lasers

νaClock Oscillator

Atoms

High-Q resonatorLaser linewidth < 1 Hz

Laser

Coherent Optical pulses out

Optical Freq. SynthesizerDivider

3) Ultrafast optical frequency comb (Clockwork)

Microwave pulses out456 986 240 494 158

Divider

Counter

Atomic ReferencesIon and neutral atoms

Optical Frequency Standardsi i i d l isensitivity and resolution

High line Q & good signal to noise ratio (stability)High line Q & good signal-to-noise ratio (stability)Δν

signal

noise

signal

δνnoise

( )τνΔ

δνNS

FWHMnoise

)(≈ τνδν 111

0⋅⋅≈

NSQnoise

, νΔν0≈Q( )τNS 0 Q

Increase S/N, or Q, by 10, decrease waiting time by 100

neutral vs. ion optical‐standards (before lattice clock)

neutral atom

ENSEMBLE of atomstrapped ion

SINGLE ionENSEMBLE of atomsin free space

SINGLE ionin Lamb-Dicke regime

, ( )REx

γλ

<< Ω << Ω⇒ Δ <<

h

Residual Doppler shiftsl f

Sideband cooling enablesLamb-Dicke regime

x λ⇒ Δ <<

Laser cooling& Doppler-free spectroscopy

Saturated-absorption;Ramsey

gRecoil & Doppler free

spectrum

&Spectroscopy

better S/N ~ (Natom)1/2 High line-Q (long τint~1/Δν)Sh t i N 1

Precision1( / ) ; /Q S N Q ν ν−× = Δ

atom

atom-atom interactionscold collision shifts

Shot noise; Nion=1

trapping EM fieldUncertainties

& -cold-collision shifts →micromotion&Improvements

Optical atomic clocks 1E-13

1E-14atio

nHg+ vs Ca (NIST, 2001)

(best) Cs fountain

Sr vs Ca

1E-15rd d

evia

︵JILA-NIST, 2008 ︶

87S 88S

1E-16stan

dar

Hg+ Q-limitYb+ vs Yb+(PTB, 2006)

Al H

Sr vs Sr(Tokyo, 2008)

1E-17

Yb vs Sr

︵NIST-JILA, 2009 ︶

Alla

n s

Sr Q-limit: N=1e4, 1sec probe time

Al+ vs Hg+

︵NIST, 2008 ︶

1 10 100 1000

Time (sec)

Sr Q limit: N 1e4, 1sec probe time

Time (sec)•Comparison with other clocks is necessary to know the stability•(Averaging time)‐1/2 dependece(Averaging time) dependece•Short time stability limited by laser spectral width

Single ion Clocks

Single Ions in Paul Traps: state‐of‐the‐art optical clocks

Very low uncertainty is possible (to 10‐18)proposed by Hans Dehmelt 1975

“Alkali like” ions

S‐D Q‐pole trans.Natural width: ~Hz

1S0‐3P0 doubly forbidden 

Natural width: ~mHz!!

“Alkali‐like” ions “Alkaline earth‐like” ions

Hg+ (NIST)Yb+ (PTB NPL)

Natural width: ~Hz… Natural width:  mHz!!

In+ (NICT?, )Al+ (NIST)Yb (PTB, NPL)

Sr+ (NPL, NRC)Ca+(Mars., Innsb., NICT)Ba+ (U Wash.)

Al+ (NIST)

( )•Strong transitions from ground states locate in VUV region→S h f li d

•Quadrupole shift→Schemes of cooling and detection needed 

•Broader natural linewidth (~Hz)  

Recent progress of ion clocks

Alkaline earth‐like ions Al+: 1S0‐3P0Natural linewidth: ~mHz

Excited state population of Al+ was effectively copied to Be+ which has a good

Insensitive to  black body radiation

P. Schmidt, Science 309 749 (2005)

Excited state population of Al+ was effectively copied to Be+ which has a good “detection” transition

Ion clocks became available for intrinsically ultranarrow transitions, and  currently NIST Al+ clock reaches 1e‐17 level

Rosenband Science 319 1808 (2008);Rosenband, Science 319, 1808 (2008);

Lattice Clocks

We can be picky! That’s a lattice clock

Single Trapped IonFree Neutral Atoms Single Trapped Ion(Accuracy)

• Tight Confinement

Free Neutral Atoms(Stability)

• Many Quantum Absorbers • Tight Confinement– No Doppler– Long Interrogation Times

• Many Quantum Absorbers– Large N

g g

• No CollisionsN∝stability

Merge together !!

Simultaneous control of induced dipolepotentials for cooling transitionpotentials for cooling transition

A strong laser light couples states connected by dipole transitionsA strong laser light couples states connected by dipole transitions.Cooling ground & excited states can be controlled independently.

5p2 3P0,1,2Singlet Triplet2

2 μωinduced polarizability

87Sr

5s5d 3D1 25s5p 1P1

5s6s 3S1

688nm

series series22

2)(

ωωω

μωωα

EEm nm

nmnmn

=−

−= ∑h

h

5s5p 3P

5s5d D1,25s5p P1 688nm λF

2.6μmλB=460nm

ω EE mnnm −=h

light shift potential 5s5p P0λF

λR=698nmCl k t iti ( ~ H )

2),()(41),( ωωαω rErU nn −=

light shift potential

5s2 1S0Clock transition(γ~mHz)4

State insensitive  Singlet states

Triplet states

optical latticeCouplingby FORT laser

λFy

3P0λFDetection of excited fraction Blue sideband

-250-200

zHkL

1S0 Elastic

-400-350-300

ktfihSHk

3P

87Sr

λ

700 750 800 850 900

-450400kratS

3P01S0

Magic λ

Sr 1S0-3P1 914nm JST, JILALaser wavelength HnmL

0 1 ,Sr 1S0-3P0 813nm Tokyo, JILA,

SYRTE, NICT

Katori et al., Phys. Rev. Lett. 91, 173005 (2003).Yb 1S0-3P1 759nm NIST, AISTHg 1S0-3P0 358nm Tokyo, SYRTE

Optical lattice clockpWorldwide spread

ストロンチウム イッテルビウム 水銀ストロンチウム イッテルビウム 水銀ストロンチウム イッテルビウム 水銀strontium ytterbium mercury

List of optical radiation to express meterList of optical radiation to express meterλ frequency uncertainty

237 nm 115In+, 5s2 1S0 - 5s5p 3P0 transition 1267402452899.92 kHz 3.6×10-13

243 nm 1H, 1S - 2S, 2 photon transition 1233030706593.55 kHz 2.0×10-13

282 nm 199Hg+, 5d106s 2S1/2 (F=0) - 5d96s2 2D5/2 (F=2) transition 1064721609899145 Hz 3×10-15

436 nm 171Yb+, 6s2S1/2 (F=0) - 5d2D3/2 (F=2) transition 688358979309308 Hz 9×10-151/2 ( ) 3/2 ( )

467 nm 171Yb+, 2S1/2 (F=0) - 2F7/2 (F=3) transition 642121496772657 Hz 6×10-14

532 nm Nd:YAG laser, 127I2, R(56)32-0:a10 563260223513 kHz 8.9×10-12

543 nm He-Ne laser 127I2 R(106)28-8:b10 551580162400 kHz 4 5×10-11543 nm He Ne laser, I2, R(106)28 8:b10 551580162400 kHz 4.5×10

578 nm 171Yb, 6s2 1S0 (F=1/2) - 6s6p 3P0 (F=1/2) transition 518295836590864 Hz 1.6×10-13

633 nm He-Ne laser, 127I2, R(127)11-5:a16 473612353604 kHz 2.1×10-11

657 nm 40Ca, 1S0 - 3P1, ΔmJ = 0 455986240494140 Hz 1.8×10-14

674 nm 88Sr+, 52S1/2 - 42D5/2 444779044095484 Hz 7×10-15

698 nm 87Sr, 5s2 1S0 - 5s5p 3P0 transition 429228004229873.65 Hz 1×10-15

698 nm 88Sr, 5s2 1S0 - 5s5p 3P0 transition 429228066418012 Hz 1×10-14

729 nm 40Ca+, 4s 2S1/2 – 3d 2D5/2 transition 411042129776393 Hz 4×10-14

778 nm 85Rb, 5S1/2(F=3) - 5D5/2(F=5), 2 photon transition 385285142375 kHz 1.3×10-11

1.5mm 13C2H2, P(16)(ν1 + ν3) transition 194369569384 kHz 2.6×10-11

3.39mm He-Ne laser, CH4, n3, P(7), F2(2) 88376181600.18 kHz 3×10-12