Detection of trapped antihydrogen in ALPHA

Detection of trapped antihydrogen in ALPHA Richard Hydomako Department of Physics and Astronomy University of Calgary WNPPC, Feb 26 2012 Image credit: Maximilien Brice, CERN

Physics motivation for antihydrogen Comparison of matter and antimatter:? Antihydrogen is the simplest antiatomic system The comparison of antihydrogen to hydrogen is a stringent CPT test 1S - 2S transition in atomic hydrogen is known to parts in 10 14 Gravitational interaction measurements 2 / 33

The ALPHA experiment ALPHA is an international experimental effort (located at the Antiproton decelerator at CERN) to produce, trap, and perform precision measurements on antihydrogen 3 / 33

The ALPHA collaboration University of Aarhus Auburn University University of British Columbia University of California, Berkeley University of Calgary CERN University of Liverpool Nuclear Research Center, Negev RIKEN Federal University of Rio de Janeiro Simon Fraser University Stockholm University Swansea University University of Tokyo TRIUMF York University G. B. Andersen, P. D. Bowe, J. S. Hangst F. Robicheaux A. Gutierrez, W. N. Hardy, S. Seif el Nasr M. Baquero-Ruiz, S. Chapman, J. Fajans, A. Povilus, J. S. Wurtele T. Friesen, R. Hydomako, R. I. Thompson E. Butler P. Nolan, P. Pusa E. Sarid D. M. Silveira, Y. Yamazaki C. L. Cesar M. D. Ashkezari, M. E. Hayden S. Jonsell W. Bertsche, M. Charlton, S. Eriksson, A. Humphries, L. V. Jørgensen, N. Madsen, D. P. van der Werf R. Hayano M. C. Fujiwara, D. R. Gill, L. Kurchaninov, K. Olchanski, A. Olin, J. W. Storey C. Amole, S. Menary 4 / 33

6 / 33

The ALPHA apparatus 7 / 33

Antihydrogen formation Positrons and antiprotons are mixed together to form antihydrogen Antiprotons are excited into the positron plasma autoresonantly (axial frequency locked to rf drive) The resulting neutral antiatom is no longer confined by the Penning trap fields 8 / 33

Trapping experiment 1. Receive 7 10 7 antiprotons from the AD and 10 7 positrons are accumulated from a radioactive 22 Na source 9 / 33

Trapping experiment 1. Receive 7 10 7 antiprotons from the AD and 10 7 positrons are accumulated from a radioactive 22 Na source 2. 10 7 Positrons and 4 10 4 antiprotons are prepared for mixing and placed in a nested-well potential 3. Neutral trap engaged 4. Particles are mixed together for 1 second, producing antihydrogen (Control measurement) antihydrogen production can be suppressed by rf-heating the positron plasma 5. Clearing pulses (500 Vm 1 ) are applied to remove charged particles 6. Neutral trap is quickly ramped down, releasing anything remaining Trapped antihydrogen will have been held for at least 172ms 9 / 33

Silicon detector 25 14 5 24 23 13 4 22 12 15 26 6 3 21 11 16 27 7 28 17 1 8 9 18 29 30 2 20 10 19 (a) At University of Liverpool (b) Module arrangement 60 double-sided silicon microstrip detector modules, arranged in three concentric layers 30 720 strips with pitch widths of 875 µm in the ẑ direction, and 227 µm in the ˆφ direction 10 / 33

Event reconstruction in the ALPHA detector Example Monte Carlo event Annihilation on the electrode surface produces several charged pions 11 / 33

Event reconstruction in the ALPHA detector The detector modules record energy deposition within the silicon wafers (in this case, by the passage of charged pions) Orthogonal signal strips give the hit positions in the plane of the silicon module 12 / 33

Event reconstruction in the ALPHA detector The charged particle tracks can be identified and extrapolated back into the apparatus Tracks are modeled as helices in an uniform magnetic field 13 / 33

Event reconstruction in the ALPHA detector The annihilation vertex is determined as the point where the tracks converge 14 / 33

Event reconstruction in the ALPHA detector The performance of the reconstruction algorithms can be evaluated using prior knowledge from the Monte Carlo simulation 15 / 33

Reconstructed vertex resolution Number of Vertices Number of Vertices 700 2 Z µ 2 χ / ndf 269.6 / 92 C 1 σ Z,1 Z,1 2 Z,1 CZ,1 372.8 ± 9.6 600 f(z) = e 2πσ2 µ Z,1 Z,1 0.004 ± 0.005 cm Z µ 2 500 C 1 σ σ Z,2 Z,2 Z,1 2 0.272 ± 0.008 cm Z,2 + e C 2πσ2 Z,2 227.5 ± 8.8 400 Z,2 µ Z,2 0.04 ± 0.03 cm 300 200 100 400 300 200 100 0 5 4 3 2 1 0 1 2 3 4 5 Z Z (cm) TRUE REC σz,2 1.291 ± 0.046 cm 500 χ2 / ndf 190.2 / 66 R µ 2 f(r) = + C 1 R,1 2 e 2πσ2 R,1 C 1 R,2 2 e 2πσ2 R,2 σ R,1 R,1 R µ σ R,2 R,2 2 CR,1 268.4 ± 16.6 µ R,1 0.0035 ± 0.0098 cm σr,1 0.318 ± 0.015 cm CR,2 343.7 ± 16.7 µ R,2 0.22 ± 0.02 cm σr,2 0.987 ± 0.029 cm The vertex position resolution can be estimated using the Monte Carlo simulation Broadening dominated by multiple scattering Example estimates from Monte Carlo: Axial resolution (top): (0.67 ± 0.04) cm Radial resolution (bottom): (0.68 ± 0.02) cm 0 5 4 3 2 1 0 1 2 3 4 5 R R (cm) TRUE REC 16 / 33

Annihilation imaging Vertex Y position (cm) 4 80 3 70 2 60 1 50 0 40 1 30 2 20 3 10 4 0 4 3 2 1 0 1 2 3 4 Vertex X position (cm) Number of Vertices Vertex distributions provide information about the physics Top left: Antihydrogen formation in the neutral-atom traop Bottom right/left: Antiproton annihilation in the octupole magnetic field Vertex Y position (cm) 4 3 2 1 0 1 2 3 100 90 80 70 60 50 40 30 20 10 Number of Vertices Number of Vertices 3500 3000 2500 2000 1500 1000 500 4 4 3 2 1 0 1 2 3 4 Vertex X position (cm) 17 / 33 0 0-10 -8-6 -4-2 0 2 4 6 8 10 Z (cm)

Cosmic ray background Cosmic muons can leave tracks through the detector (top right) Unsuppressed rate of 10 event/s Need to discriminate between signal and background events, especially for the detection of rare events Focused on the topological differences between cosmic and annihilation events 18 / 33 (c) Example cosmic event (d) Example annihilation

Discriminating variable (linear residual, δ) Annihiltion event Cosmic ray event Cosmic events will conform to a straight line fit Slight curvature in the strong axial magnetic field 19 / 33

Discriminating variable (vertex radius, R) Annihiltion event Cosmic ray event Cosmic events are unconstrained in radius, while annihilations occur withing the trap region of the apparatus 20 / 33

Background rejection analysis Analysis focused on detecting rare events during the trapping experiments Studies performed on auxiliary datasets to avoid optimizing on the trapping events (blind analysis): Signal: antihydrogen annihilation during mixing Background: apparatus operating without antiparticles Find cuts that optimize the expected signal 21 / 33

2.5 16.5 Cut optimization Cut on the linear residual, δ cut (cm 2 ) 1.0 0.8 0.6 0.4 0.2 5.0 10.0 14.0 16.0 N tracks > 2 0.0 0 1 2 3 4 5 6 7 8 Cut on the vertex radial position, R cut (cm) 16 14 12 10 8 6 4 2 0 14.0 Expected signal significance (σ) Performed 5000 pseudoexperiments for each cut configuration Figure of merit, the expected p-value: b n e b α =, n=n 0 n! n 0 : expected number of events b: background rate 22 / 33

2.5 5.0 16.6 Cut optimization 5 N tracks = 2 16 Cut on the linear residual, δ cut (cm 2 ) 4 3 2 1 10.0 14.0 16.0 16.5 14 12 10 8 6 4 2 Expected signal significance (σ) N tracks = 2 events are considered separately White crosses indicate the cut choices 0 0 1 2 3 4 5 6 7 8 Cut on the vertex radial position, R cut (cm) 0 5.0 23 / 33

Cosmic background rejection results N tracks Vertex radius, R cut (cm) Linear residual, δ cut (cm 2 ) = 2 < 4 > 2 > 2 < 4 > 0.05 Table: Final background rejection cuts. Events that satisfy these conditions are accepted as signal. Results of the cut optimization: (99.55 ± 0.02)% cosmic background rejection (64.4 ± 0.1)% signal acceptance (47 ± 2) 10 3 event/s background acceptance rate 24 / 33

Summary of 2010 experiments Type Number of cycles Vertices passing all cuts Normal trapping experiments 335 48 Heated positron plasma 246 1 For readouts less than 30 ms after neutral-trap magnets shutdown 25 / 33

Summary of 2010 experiments Type Number of cycles Vertices passing all cuts No bias 137 20 Left bias 101 14 Right bias 97 14 No bias, heated positrons 132 1 Left bias, heated positrons 60 0 Right bias, heated positrons 54 0 For readouts less than 30 ms after neutral-trap magnets shutdown 26 / 33

Experimental observation of trapped antihydrogen 50 a) Time (ms) 40 30 20 10 0 50 b) 40 30 20 10 0 Time-cut 20 10 0 10 20 Distance along trap axis, z (cm) a) Simulated antihydrogen signal in grey b) Coloured dots are simulated bare antiprotons 27 / 33

Experimental observation of trapped antihydrogen 50 a) Time (ms) 40 30 20 10 0 50 b) 40 30 20 10 0 Time-cut 20 10 0 10 20 Distance along trap axis, z (cm) No bias Left bias Right bias Heated e + no bias Heated e + right bias a) Simulated antihydrogen signal in grey b) Coloured dots are simulated bare antiprotons 28 / 33

Extended confinement a 2.5 b Trappingrateper attempt Significance( ) 2.0 1.5 1.0 0.5 0 10.0 7.5 5.0 2.5 0 500 1,000 1,500 2,000 Confinement time(s) 0 0 500 1,000 1,500 2,000 Confinement time(s) Neutral trap engaged for longer Trapped atoms for up to 1000 s (low significance at 2000 s) More than enough time for the antihydrogen atom to radiate to the ground state 29 / 33

What s next for ALPHA 2011 Run: introduced microwaves into the apparatus Currently: a new apparatus with laser access is being constructed 30 / 33

Conclusion ALPHA has successfully demonstrated the magnetic confinement of antihydrogen for as long as 1000 s The silicon detector and annihilation reconstruction routines played a large role in this effort The cosmic background rejection analysis enabled a sensitive identification of annihilation events The position-sensitive reconstruction allowed for discrimination against mirror-trapped antiprotons and cosmic-ray muons Long-time magnetic confinement of antihydrogen atoms will allow for precision studies of this anti-atomic system 31 / 33

Backup slides 32 / 33

Intentially mirror-trapped antiprotons 50 40 No bias Left bias Right bias Time (ms) 30 20 10 0 20 10 0 10 20 Distance along trap axis, z (cm) 33 / 33

Penning-Malmberg trap for charged particles/plasmas Strong external solenoidal magnetic field for radial confinement Electric potential for axial confinement and manipulation Provides excellent confinement of non-neutral plasmas 34 / 33

Radial compression using the Rotating Wall technique Rotating wall waveform generator 35 / 33 phase shift 0 90 180 270 Segmented electrode y x Rotating electric field applies a torque to the rotating plasma The mean-squared plasma radius is related the canonical angular momentum (e. g. electron plasma): P θ = ( eb/2c) r 2 An applied torque T = dp θ /dt can then be used to increase or decrease the plasma radius The rotating wall is typically driven as a dipole with 0.5-2.5 V, 0.5-20 MHz, with optional sweeping frequency

Evaporative cooling of charged plasmas Demonstrated evaporative cooling of antiprotons to temperatures as low as 9 K Energetic particles escape as the confining potential is lowered, leaving the remaining particles at a lower temperature Also can be applied to the positrons plasma, resulting in a positron temperature of about 40 K G. B. Andresen et al. (ALPHA Collaboration), Phys. Rev. Lett. 105, 013003 (2010). 36 / 33

Autoresonant excitation of the antiproton plasma Response of a driven nonlinear oscillator (J. Fajans and L. Friedland, Am. J. Phys. 69, 1096 (2001)) 5 (a) (b) 4 3 2 1 0 0.7 0.8 0.9 1.0 Response of a driven antiproton plasma 37 / 33 The electrostatic confining wells we use have anharmonic components The oscillation amplitude (parallel energy) is a function of the bounce frequency Equation of motion: θ + ω 2 0 sin θ = ɛ cos(ω it αt 2 /2) A drive with a decreasing frequency can result in an increase in the oscillation amplitude Typical drive: 200 µs, 55 mv, 350-200 khz

Magnetic neutral-atom trap Antihydrogen has a small magnetic moment, which interacts with the field, U = µ B The ALPHA neutral trap consists of a superconducting octupole (left) for the radially increasing field, and two mirror coils for the axial field Shallow trap depth: 0.5 K 38 / 33