Gabrielse

New Measurement
of the Electron Magnetic Moment
and the Fine Structure Constant
Gerald Gabrielse
Leverett Professor of Physics
Harvard University
Almost finished student: David Hanneke 2006 DAMOP Thesis
Earlier contributions: Brian Odom, Prize Winner
Brian D’Urso,
Steve Peil, 
2
20 years
Dafna Enzer,
6.5 theses
Kamal Abdullah
Ching-hua Tseng
Joseph Tan
N$F 0.1 m
 Gabrielse

Recent Back-to-Back Papers

New Measurement of the Electron Magnetic Moment
B. Odom, D. Hanneke, B. D’Urson and G. Gabrielse,
Phys. Rev. Lett. 97, 030801 (2006).
New Determination of the Fine Structure Constant
G. Gabrielse, D. Hanneke, T. Kinoshita, M. Nio, B. Odom,
Phys. Rev. Lett. 97, 030802 (2006).

AIP Physics Story of the Year (Phys. News Update, 5 Dec. 2006)
• Science 313, 448-449 (2006)
• Nature 442, 516-517 (2006)
• Physics Today, 15-17 (August, 2006)
• Cern Courier (October 2006)
• New Scientist 2568, 40-43 (2006)
• Physics World (March 2007)
 Gabrielse
Why Does it take Twenty Years and 6.5 Theses?
Explanation 1: Van Dyck, Schwinberg, Dehemelt did a good job in 1987!
Phys. Rev. Lett. 59, 26 (1987)
Explanation 2a: We do experiments much too slowly
Explanation 2b: Takes time to develop new ideas and methods
needed to measure with 7.6 parts in 1013 uncertainty
• One-electron quantum cyclotron
first measurement with

• Resolve lowest cyclotron states as well as spin

these new methods

• Quantum jump spectroscopy of spin and cyclotron motions

• Cavity-controlled spontaneous emission
• Radiation field controlled by cylindrical trap cavity
• Cooling away of blackbody photons
• Synchronized electrons identify cavity radiation modes
• Trap without nuclear paramagnetism
• One-particle self-excited oscillator
 Gabrielse
The New Measurement of Electron g

U. Michigan U. Washington Harvard

beam of electrons one electron one electron

spins precess observe spin quantum

with respect to flip cyclotron 100 mK
cyclotron motion motion
thermal
cyclotron resolve lowest self-excited
motion quantum levels oscillator

cavity-controlled inhibit spontan.

radiation field emission
Dehmelt, (cylindrical trap)
Crane, Rich, … Van Dyck cavity shifts
 Gabrielse
Magnetic Moments, Motivation and Results
 Gabrielse
Magnetic Moments

magnetic  L angular momentum
  g B
moment 
Bohr magneton e
2m

e.g. What is g for identical charge and mass distributions?

v e, m
e ev  L e e L 
  IA  ( 2 )   L
 2  2 mv 2m 2m 
 
 v 
 g 1 B
 Gabrielse
Magnetic Moments

magnetic  S angular momentum
  g B
moment 
Bohr magneton e
2m

g  1 identical charge and mass distribution

g2 spin for Dirac point particle

g  2.002 319 304 ... simplest Dirac spin, plus QED

(if electron g is different  electron has substructure)

 Gabrielse
Why Measure the Electron Magnetic Moment?
1. Electron g - basic property of simplest of elementary particles
2. Determine fine structure constant – from measured g and QED
(May be even more important when we change mass standards)
3. Test QED – requires independent 
4. Test CPT – compare g for electron and positron  best lepton
test
5. Look for new physics beyond the standard model
• Is g given by Dirac + QED? If not  electron substructure
(new physics)
• Muon g search needs electron g measurement
 Gabrielse
New Measurement of Electron Magnetic Moment

magnetic  S spin
  g B
moment 
Bohr magneton e
2m

g / 2  1.001 159 652 180 85

 0.000 000 000 000 76 7.6 1013

• First improved measurement since 1987

• Nearly six times smaller uncertainty
• 1.7 standard deviation shift
• Likely more accuracy coming
• 1000 times smaller uncertainty than muon g
B. Odom, D. Hanneke, B. D’Urso and G. Gabrielse,
Phys. Rev. Lett. 97, 030801 (2006).
 Gabrielse

0 85 (76)

(more digits coming)

 Gabrielse

Dirac + QED Relates Measured g and Measured 

2 3 4
g        
 1  C1    C2    C3    C4    ...  a
2        
Dirac
weak/strong
point
particle
Sensitivity to other physics
Measure QED Calculation (weak, strong, new) is low
Kinoshita, Nio,
Remiddi, Laporta, etc.

1. Use measured g and QED to extract fine structure constant

2. Wait for another accurate measurement of Test QED
 Basking in the Reflected Glow of TheoristsGabrielse
g  
 1  C1  
2  
2
 
 C2  
 
3
 
 C3  
 
4
 
 C4  
 
5
 
 C5   2004
 
Remiddi Kinoshita G.G
 ...  a
 Gabrielse
2 3 4
g        
 1  C1    C2    C3    C4    ...  a
2        
theoretical uncertainties

experimental
uncertainty
 Gabrielse
New Determination of the Fine Structure Constant
1 e 2 • Strength of the electromagnetic interaction
 • Important component of our system of
4 0 c fundamental constants
• Increased importance for new mass standard
 1  137.035 999 710
 0.000 000 096 7.0 1010

• First lower uncertainty

since 1987
• Ten times more accurate than
atom-recoil methods

G. Gabrielse, D. Hanneke, T. Kinoshita, M. Nio, B. Odom,

Phys. Rev. Lett. 97}, 030802 (2006).
 Gabrielse
Widely Re-Reported

Science
Nature
Physics Today
New Scientist
Cern Courier
…
Fox News

Moral: what is quoted

is not necessarily
what was said
 Gabrielse
Next Most Accurate Way to Determine use Cs example)
Combination of measured Rydberg, mass ratios, and atom recoil
1 e2 1 e 4 me c
   R  Haensch, …
4 0 hc (4 0 ) 2 2h3c 2
2 R h
  2
Pritchard, …
c me Chu, …
2 R h M Cs M p h 2 f recoil
   2c
c M Cs M p me M Cs ( f D1 )2
Haensch, …
f recoil M Cs M 12C Tanner, …
 2  4 R c
( f D1 ) 2 M 12C me
Werthe, Quint, … (also Van Dyck)
Biraben, …

• Now this method is 10 times less accurate

• We hope that will improve in the future  test QED

(Rb measurement is similar except get h/M[Rb] a bit differently)

 Gabrielse
Earlier Measurements
Require Larger Uncertainty Scale

ten times
larger scale
to see larger
uncertainties
 Gabrielse
Test of QED

Most stringent test of QED: Comparing the measured electron g

to the g calculated from QED using
an independent 

 g  15  10 12

• The uncertainty does not comes from g and QED

• All uncertainty comes from [Rb] and [Cs]
• With a better independent  could do a ten times better test
 Gabrielse
From Freeman Dyson – One Inventor of QED
Dear Jerry,
... I love your way of doing experiments, and I am happy to congratulate you for
this latest triumph.  Thank you for sending the two papers.
Your statement, that QED is tested far more stringently than its inventors could
ever have envisioned, is correct.  As one of the inventors, I remember that we
thought of QED in 1949 as a temporary and jerry-built structure, with
mathematical inconsistencies and renormalized infinities swept under the rug.  We
did not expect it to last more than ten years before some more solidly built theory
would replace it.  We expected and hoped that some new experiments would
reveal discrepancies that would point the way to a better theory. And now, 57
years have gone by and that ramshackle structure still stands. The theorists …
have kept pace with your experiments, pushing their calculations to higher
accuracy than we ever imagined. And you still did not find the discrepancy that
we hoped for.  To me it remains perpetually amazing that Nature dances to the
tune that we scribbled so carelessly 57 years ago.  And it is amazing that you can
measure her dance to one part per trillion and find her still following our beat.
With congratulations and good wishes for more such beautiful experiments, yours
ever, Freeman.
 Gabrielse
Direct Test for Physics Beyond the Standard Model

g  2  2aQED ( )   g SM :Hadronic Weak   g New Physics

Is g given by Dirac + QED? If not  electron substructure

Does the electron have internal structure? Brodsky, Drell, 1980

m limited by the uncertainty in
m*   130 GeV / c 2
g/2 independent values
m
m*   600 GeV / c 2 if our g uncertainty
g/2
was the only limit
Not bad for an experiment done at 100 mK, but LEP does better
m*  10.3 TeV LEP contact interaction limit
 Gabrielse
Muon Test for Physics Beyond the Standard Model
Needs Measured Electron g

less accurately measured expected to be bigger

than we measure electron g than for electron
by a factor of 1000 by ~40,000

g  2  2aQED ( )   g SM :Hadronic Weak   g New Physics

big contribution need 

must be subtracted out need test the QED calculation
of this large contribution

 Muon search for new physics

needs the measurement of the electron g and 
 Gabrielse
Could We Check the 3 Disagreement
between Muon g Measurement and “Calculation”?

g  2  2aQED ( )   g SM :Hadronic Weak   g New Physics

(m/me)2 ~ 40,000  muon more sensitive to “new physics”
÷1,000  how much more accurately we measure
÷ 3  3 disagreement is now seen
 If we can reduce the electron g uncertainty by 13 times more
should be able to have the precision to see the 3 effect (or not)

Also need: • QED and SM calculations improved by factor of ~5

• Independent measurement of  improved by factor of 130

These are large numbers  hard to imagine that this will happen quickly
 Gabrielse

How Does One Measure the Electron g

to 7.6 parts in 1013?
 Gabrielse
How to Get an Uncertainty of 7.6 parts in 1013

• One-electron quantum cyclotron

first measurement with

• Resolve lowest cyclotron as well as spin states

these new methods

• Quantum jump spectroscopy of cyclotron and spin motions

• Cavity-controlled spontaneous emission
• Radiation field controlled by cylindrical trap cavity
• Cooling away of blackbody photons
• Synchronized electrons probe cavity radiation modes
• Elimination of nuclear paramagnetism
• One-particle self-excited oscillator

Make a “Fully Quantum Atom” for the electron

Challenge: An elementary particle has no internal states to

probe or laser-cool

 Give introduction to some of the new and novel methods

 Gabrielse
Basic Idea of the Measurement

Quantum jump spectroscopy

of lowest cyclotron and spin levels
of an electron in a magnetic field
 Gabrielse
One Electron in a Magnetic Field

 c  150 GHz 2


n=4
n=3
n=2 0.1
n=1 hc  7.2 kelvin m
n=0
2


Need low
B  6 Tesla temperature
cyclotron motion
T << 7.2 K 0.1
m
 Gabrielse
First Penning Trap Below 4 K  70 mK
Need low
temperature
cyclotron motion
T << 7.2 K
 Gabrielse

David Hanneke G.G.

 Gabrielse

Electron Cyclotron Motion

Comes Into Thermal Equilibrium

T = 100 mK << 7.2 K  ground state always

Prob = 0.99999…

cold electron
hot
cavity blackbody
spontaneous
emission photons
 Gabrielse
Electron in Cyclotron Ground State
QND Measurement of Cyclotron Energy vs. Time

0.23

0.11

0.03

9 x 10-39

average number On a short time scale

of blackbody  in one Fock state or another
photons in the Averaged over hours
cavity  in a thermal state
S. Peil and G. Gabrielse, Phys. Rev. Lett. 83, 1287 (1999).
 Gabrielse
Spin  Two Cyclotron Ladders of Energy Levels

n=4
c
n=3
n=4 c c
Cyclotron n=2 Spin
n=3 c c
frequency: n=1 frequency:
n=2 c c
1 eB n=0 g
c  n=1 c s  c
2 m 2
n=0
ms = -1/2 ms = 1/2
 Gabrielse
Basic Idea of the Fully-Quantum Measurement

n=4
c
n=3
n=4 c c
Cyclotron n=2 Spin
n=3 c c
frequency: n=1 frequency:
n=2 c c
1 eB n=0 g
c  n=1 c s  c
2 m 2
n=0
ms = -1/2 ms = 1/2

g s  s  c B in free
Measure a ratio of frequencies:   1
2 c c space
103
• almost nothing can be measured better than a frequency
• the magnetic field cancels out (self-magnetometer)
 Gabrielse
Special Relativity Shift the Energy Levels 

n=4
 c  9 / 2
n=3
n=4  c  7 / 2  c  7 / 2
Cyclotron n=2 Spin
n=3  c  5 / 2  c  5 / 2
frequency: n=1 frequency:
n=2  c  3 / 2  c  3 / 2
eB n=0 g
2 c  n=1 c  / 2 s  c
m 2
n=0
ms = -1/2 ms = 1/2

Not a huge relativistic shift,  h c 9

  10
but important at our accuracy  c mc 2

Solution: Simply correct for  if we fully resolve the levels

(superposition of cyclotron levels would be a big problem)
 Gabrielse
Cylindrical Penning Trap
V 2 z 2  x 2  y 2

• Electrostatic quadrupole potential  good near trap center

• Control the radiation field  inhibit spontaneous emission by 200x
(Invented for this purpose: G.G. and F. C. MacKintosh; Int. J. Mass Spec. Ion Proc. 57, 1 (1984)
 Gabrielse
One Electron in a Penning Trap
• very small accelerator
• designer atom

cool 12 kHz 200 MHz detect

need to
Electrostatic 153 GHz measure
quadrupole for g/2
Magnetic field
potential
 Gabrielse
Frequencies Shift
Imperfect Trap
• tilted B
Perfect Electrostatic • harmonic
B in Free Space Quadrupole Trap distortions to V
eB
c   c '  c c
m
 z  c ' z
 m  z m
g g g
s  c s  c s  c
2 2 2
g s not a measurable eigenfrequency in an
Problem: 
2 c imperfect Penning trap
Solution: Brown-Gabrielse invariance theorem
 c  ( c )2  ( z ) 2  ( m ) 2
 Gabrielse

Spectroscopy in an Imperfect Trap

• one electron in a Penning trap
• lowest cyclotron and spin states

g  s vc  ( s  c ) vc  a
  
2 c c c
( z ) 2
a 
g 2 c
 1
2 3 ( z ) 2
fc  
2 2 c
expansion for vc  z  m 

To deduce g  measure only three eigenfrequencies

of the imperfect trap
 Gabrielse
Detecting and Damping Axial Motion
measure voltage

V(t)

I2R
damping
Axial motion
200 MHz
of
trapped
electron self-excited
oscillator
feedback
amplitude
 Gabrielse
Feedback Cooling of an Oscillator
Electronic Amplifier Feedback: Strutt and Van der Ziel (1942)
Basic Ideas of Noiseless Feedback and Its Limitations: Kittel (1958)
Dissipation :  e  (1  g ) Fluctuations: Te  T (1  g ) faster damping rate
Fluctuation-Dissipation Invariant: e / Te  const  higher temperature

Applications: Milatz, … (1953) -- electrometer

Dicke, … (1964) -- torsion balance
Forward, … (1979) -- gravity gradiometer
Ritter, … (1988) -- laboratory rotor
Cohadon, … (1999) -- vibration mode of a mirror
Proposal to apply Kittel ideas to ion in an rf trap
Dehmelt, Nagourney, … (1986)  never realized
Proposal to “stochastically” cool antiprotons in trap
Beverini, … (1988) – stochastic cooling  never realized
Rolston, Gabrielse (1988) – same as feedback cooling (same limitations)

Realization of feedback cooling with a trapped electron (also include noise)

D’Urso, Odom, Gabrielse, PRL (2003)
 Gabrielse
one-electron self-excited oscillator QND Detection
of One-Quantum Transitions
 1
B B2 z 2  H  m z 2 z 2   B2 z 2
2

n=0 n=1 n=0

cyclotron cyclotron cyclotron
ground excited ground
state state state

n=1
freq Ecyclotron  hf c (n  12 )
n=0
time
 QND Gabrielse

Quantum Non-demolition Measurement

H = Hcyclotron + Haxial + Hcoupling

[ Hcyclotron, Hcoupling ] = 0 QND

condition
QND: Subsequent time evolution
of cyclotron motion is not
altered by additional
QND measurements
 Observe Tiny Shifts of the Frequency Gabrielse
of a One-Electron Self-Excited Oscillator

one quantum
cyclotron
excitation

spin flip

Unmistakable changes in the axial frequency

signal one quantum changes in cyclotron excitation and spin

"Single-Particle Self-excited Oscillator"

B B. D'Urso, R. Van Handel, B. Odom and G. Gabrielse
Phys. Rev. Lett. 94, 113002 (2005).
 Gabrielse

Emboldened by the Great Signal-to-Noise

Make a one proton (antiproton) self-excited oscillator

 try to detect a proton (and antiproton) spin flip
• Hard: nuclear magneton is 500 times smaller
• Experiment underway  Harvard
 also Mainz and GSI (without SEO)
(build upon bound electron g values)
 measure proton spin frequency
 we already accurately measure antiproton cyclotron frequencies
 get antiproton g value (Improve by factor of a million or more)
 Gabrielse
Need Averaging Time to Observe
a One-quantum Transition
 Cavity-Inhibited Spontaneous Emission
Application of Cavity QED

excite,
number of n=1 to n=0 decays

30
measure time in excited state
 = 16 s
20

10

0 10 20 30 40 50 60

decay time (s)

 Gabrielse
Cavity-Inhibited Spontaneous Emission

1
Free Space  
75 ms
B = 5.3 T

Within 1
  Inhibited
Trap Cavity 16 sec By 210!
B = 5.3 T

cavity
modes
Purcell
Kleppner
c frequency
Gabrielse and Dehmelt
 Gabrielse
“In the Dark” Excitation  Narrower Lines
1. Turn FET amplifier off
2. Apply a microwave drive pulse of ~150 GH
(i.e. measure “in the dark”)
3. Turn FET amplifier on and check for axial frequency shift
4. Plot a histograms of excitations vs. frequency
# of cyclotron excitations

Good amp heat sinking,

amp off during excitation
Tz = 0.32 K
0 100 200 300

frequency - c (ppb)
 Gabrielse
Big Challenge: Magnetic Field Stability
Magnetic field cancels out
n=2
n=3
n=1 g s a
n=2   1
n=1 n=0 2 c c
n=0 ms = 1/2
ms = -1/2 But: problem when B
drifts during the
measurement

Magnetic field take

~ month to stabilize
 Gabrielse
Self-Shielding Solenoid Helps a Lot
Flux conservation  Field conservation
Reduces field fluctuations by about a factor > 150

“Self-shielding Superconducting Solenoid Systems”,

G. Gabrielse and J. Tan, J. Appl. Phys. 63, 5143 (1988)
 Gabrielse
Eliminate Nuclear Paramagnetism
Deadly nuclear magnetism of copper and other “friendly” materials
 Had to build new trap out of silver ~ 1 year
 New vacuum enclosure out of titanium setback
Gabrielse
Gabrielse
 Gabrielse

Quantum Jump Spectroscopy

• one electron in a Penning trap
• lowest cyclotron and spin states
 Gabrielse
Measurement Cycle

n=3 n=2
g s a n=2 n=1
  1 n=0
2 c c n=1
n=0 ms = 1/2
simplified ms = -1/2

1. Prepare n=0, m=1/2  measure anomaly transition

3 hours 2. Prepare n=0, m=1/2  measure cyclotron transition

0.75 hour 3. Measure relative magnetic field

Repeat during magnetically quiet times

 Gabrielse
Measured Line Shapes for g-value Measurement
It all comes together:
• Low temperature, and high frequency make narrow line shapes
• A highly stable field allows us to map these lines

cyclotron anomaly

n=3 n=2
n=2 n=1
n=1 n=0
n=0 ms = 1/2
ms = -1/2

Precision:
Sub-ppb line splitting (i.e. sub-ppb precision of a g-2 measurement)
is now “easy” after years of work
 Gabrielse
Cavity Shifts of the Cyclotron Frequency

n=3 n=2
g s a n=1
  1 n=2
2 c c n=1 n=0
n=0 ms = 1/2
ms = -1/2

1 spontaneous emission
 
16 sec inhibited by 210
B = 5.3 T
Within a Trap Cavity

cavity cyclotron frequency

modes is shifted by interaction
with cavity modes
c frequency
 Gabrielse
Cavity modes and Magnetic Moment Error
use synchronization of electrons to get cavity modes

Operating between modes of cylindrical trap first measured

where shift from two cavity modes cavity shift of g
cancels approximately
 Gabrielse
Summary of Uncertainties for g (in ppt = 10-12)
Test of
cavity
shift Measurement
understanding of g-value
Gabrielse
 Gabrielse
Attempting to Measure g for Proton and Antiproton
• Improve proton g by more than 10
• Improve antiproton g by more than 106

• Compare g for antiproton and proton – test CPT

 Gabrielse
Current Proton g Last Measured in 1972
CODATA 2002: gp=5.585 694 701(56) (10 ppb)

 p  H  ge  H  g p m p
g p  ge
e  H  ge g p ( H ) me
proton-electron mass ratio,
measured to < 1 ppb
(Mainz)
electron g-factor, bound / free corrections,
measured to calculated to < 1 ppb
(Breit, Lamb, Lieb, Grotch, Faustov,
< 0.001 ppb Close, Osborn, Hegstrom, Persson,
(Harvard)
others)

ge  H  1 1 1 2   1 2 m 
 1   Z    Z    Z       Z    e   
2 4
bound magnetic moment ratio, ge 3 12 4   2  mp 
 
measured to 10 ppb
(MIT: P.F. Winkler, D. Kleppner,  1  17.7053 106
T. Myint, F.G. Walther, 
Phys. Rev. A 5, 83-114 (1972) )
gp  H  1 1  m   3  4a p 
 1  Z 2  Z 2  e   
gp 3 6  m   1  a 
 p  p 

 1  17.7328 10 6
 Gabrielse
History of Measurements of Proton g

(from bound measurements of p/e,

with current values of ge, me/mp and theory)
 Gabrielse
Antiproton g-factor

Antiproton g-factor is known to less than a part per thousand

g p  5.60118

We hope to do roughly one million times better.

 Gabrielse
Apparatus Working Only With Electrons (so far)

iron detect spin

flip

make spin
flip

6 mm inner
diameter
Nick Guise
 Gabrielse
Summary and Conclusion
 Gabrielse

Summary

How Does One Measure g to 7.6 Parts in 1013?

 Use New Methods

• One-electron quantum cyclotron
first measurement with

• Resolve lowest cyclotron as well as spin states

these new methods

• Quantum jump spectroscopy of lowest quantum states

• Cavity-controlled spontaneous emission
• Radiation field controlled by cylindrical trap cavity
• Cooling away of blackbody photons
• Synchronized electrons probe cavity radiation modes
• Trap without nuclear paramagnetism
• One-particle self-excited oscillator
 Gabrielse
New Measurement of Electron Magnetic Moment

magnetic  S spin
  g B
moment 
Bohr magneton e
2m

g / 2  1.001 159 652 180 85

 0.000 000 000 000 76 7.6 1013

• First improved measurement since 1987

• Nearly six times smaller uncertainty
• 1.7 standard deviation shift
• Likely more accuracy coming
• 1000 times smaller uncertainty than muon g
B. Odom, D. Hanneke, B. D’Urso and G. Gabrielse,
Phys. Rev. Lett. 97, 030801 (2006).
 Gabrielse
New Determination of the Fine Structure Constant
1 e 2 • Strength of the electromagnetic interaction
 • Important component of our system of
4 0 c fundamental constants
• Increased importance for new mass standard
 1  137.035 999 710
 0.000 000 096 7.0 1010

• First lower uncertainty

since 1987
• Ten times more accurate than
atom-recoil methods

G. Gabrielse, D. Hanneke, T. Kinoshita, M. Nio, B. Odom,

Phys. Rev. Lett. 97}, 030802 (2006).
 Gabrielse
We Intend to do Better
Stay Tuned – The new methods have just been made to work
all together
• With time we can utilize them better
• Some new ideas are being tried (e.g. cavity-sideband cooling)
• Lowering uncertainty by factor of 13  check muon result (hard)

Spin-off Experiments
• Use self-excited antiproton oscillator to measure the
antiproton magnetic moment  million-fold improvement?

• Compare positron and electron g-values to make best test

of CPT for leptons

• Measure the proton-to-electron mass ration directly

 Gabrielse

Further Reading
New Measurement of the Electron Magnetic Moment
B. Odom, D. Hanneke, B. D’Urson and G. Gabrielse,
Phys. Rev. Lett. 97, 030801 (2006).
New Determination of the Fine Structure Constant
G. Gabrielse, D. Hanneke, T. Kinoshita, M. Nio, B. Odom,
Phys. Rev. Lett. 97, 030802 (2006).

AIP Physics Story of the Year (Phys. News Update, 5 Dec. 2006)
• Science 313, 448-449 (2006)
• Nature 442, 516-517 (2006)
• Physics Today, 15-17 (August, 2006)
• Cern Courier (October 2006)
• New Scientist 2568, 40-43 (2006)
• Physics World (March 2007)