1. 1/37 April 2007Walter Scandale H8-RD22 Experiment: progress on ion beam focusing with bent crystals Walter Scandale CERN For the H8-RD22 collaboration (CERN, FNAL, INFN, IHEP, JINR, PNPI) ALICE seminar 30 April 2007 We acknowledge the support of the European Community-Research Infrastructure Activity under the FP6 “Structuring the European Research Area” programme (CARE, contract number RII3-CT-2003-506395).

2. 2/37 April 2007Walter Scandale Outlook  Reminder of the concept of collimation  Why using crystals  The H8-RD22 experiment at CERN  Silicon crystals  Experimental layout  High precision goniometric system  Tracking detectors  The results of the 2006 run  Crystal Angular Scans (Strip and Quasi-Mosaic Crystals)  Double Reflection Effect  Concluding remarks and future plans

3. 3/37 April 2007Walter Scandale Secondary halo p p e  Primary collimator (scatterer) Beam Core Shower Beam propagation Impact parameter ≤ 1  m Sensitive equipment Primary halo (p) e  Shower p Tertiary halo Secondary collimator (massive absorber) Two stage collimation

4. 4/37 April 2007Walter Scandale Two stage collimation phase space  short primary scatterer  longer secondary collimator downstream with larger gap aperture  halo particles captured through amplitude increase via multiple scattering and multi-turn effect. Primary collimator (scatterer) Secondary collimator (conversion in hadr. shower )  x’ x’ x  2 N 1 N Capture condition:  x’ mainly due to multiple Coulomb scattering, with ~ L Curtesy of Bellodi

5. 5/37 April 2007Walter Scandale Requirements for LHC Courtesy of R. Assmann

6. 6/37 April 2007Walter Scandale IR3 and IR7 insertions are equipped with 54 collimators made of carbon-carbon Open problems: u Resistive impedance (up to 100 times the whole LHC) u Electron cloud (local concentration) Collimation aperture Courtesy of R. Assmann

7. 7/37 April 2007Walter Scandale All the machine All the machine with Cu coated (5  m) collimators Without collimators (TCDQ+RW+BB) LHC stability diagram Courtesy of E. Metral

8. 8/37 April 2007Walter Scandale Ion collimation : why an issue? Nominal ion beam has 100 times less beam power than proton beam, but particle-collimator physics very different: ~20 times higher probability of nuclear interactions respect to p High probability to undergo nuclear interactions in the scatterer before 2-stage collimation condition is satisfied A new disturbance respect to p Curtesy of Bellodi Large variety of daughter nuclei, Monte Carlo calculated specific x-sections loss 1 n (59%)  207 Pb or 2 n (11%)  206 Pb

9. 9/37 April 2007Walter Scandale Ion collimation : what can fail? Curtesy of Bellodi 204 Pb -1.92% 205 Pb -1.44% 206 Pb -0.96% 207 Pb -0.48% 208 Pb 0.0% 203 Tl -1.2% 204 Tl -0.71% 205 Tl -0.23% 206 Tl 0.26% 207 Tl 0.75% 202 Hg -0.46% 203 Hg 0.04% 204 Hg 0.53% 205 Hg 1.02% 206 Hg 1.51% Typical transverse momentum transferred in NF ~ 1 A∙MeV/c, even smaller for ED processes (compared to ~10 A∙MeV/c due to beam emittance). After first impact/grazing with scatterer:  high probability of nuclear interactions  production of isotopes with different Z/A ratio and momentum and direction almost unchanged  secondary ions not intercepted by the secondary and lost downstream in SC magnets because of different Br LHC energy acceptance: - arcs: ~ ±1% - IR3: ~ ±0.2%

10. 10/37 April 2007Walter Scandale Loss map from computer simulations Curtesy of Bellodi Nominal ion beam with collision optics and standard collimator settings: ~50% current limit on nominal 208 Pb ion beam (due to collimation inefficiency) not to be taken at face value:  X-sections for nuclear processes have estimated errors of ~±50%  8.5 W/m heat load limit is only an average from early studies (Jeanneret et al.1996). Should really be considered on a magnet-to-magnet basis   coll ~4% depends on impact parameter distribution on the scatterer (input assumption, difficult to predict..)  12 min lifetime is another arbitrary number.

11. 11/37 April 2007Walter Scandale p Beam propagation Sensitive equipment Primary halo (p) Absorber  Primary halo directly extracted!  Much less secondary and tertiary halos!? e  Shower Crystal Crystal collimation Beam Core..but no enough data available to substantiate the idea…

12. 12/37 April 2007Walter Scandale  A bent crystal deflects halo particles toward a downstream absorber: the selective and coherent scattering the selective and coherent scattering on atomic planes of an aligned Si-crystal may replace more efficiently random scattering the random scattering process on single atoms of an amorphous scatterer. Crystal collimation: a smart approach for primary collimation amorphous scatterer Larger collimation efficiency Larger gap of the secondary collimator --> reduced impedance The hope is to get

13. 13/37 April 2007Walter Scandale The RD22 Collaboration, CERN DRDC 94-11  Large channeling efficiency measured for the first time  Consistent with simulation expectation extended to high energy beams  Experimental proof of multi-turn effect (channeling after multi-traversals)  Definition of a reliable procedure to measure the channeling efficiency RD 22: extraction of 120 GeV protons (SPS: 1990-95)

14. 14/37 April 2007Walter Scandale RD 22: extraction of ion (SPS: 1996-97)  Extraction experiments on bent crystals channelling were conducted at the SPS with ion beams in 1996-1997 (G.Arduini, W.Herr, J.Klem, L.Gatignon et al.)  Direct-type experiment on fully stripped Pb 82+ ions at 400 GeV/c/u (external beamline) and 270 GeV/c/u (stored beam in ring with different excitation levels)  60×18 mm crystal, bent over 50mm, 4mrad deflection, 3-points bend Theory: Critical angle, dechannelling length and MS expected to depend on p/Z ratio  ions should behave like p of equivalent momentum per charge Nuclear interaction rate expected to be strongly decreased for channelled vs non-channelled ions EMD should also decrease by a factor 10-100 ….. however not confirmed by measurements ….

15. 15/37 April 2007Walter Scandale RD 22: extraction of ions (SPS: 1996-97)  Single pass experiment – external beamline Very good agreement with theoretical model, corroborates expectations Ion channelling demonstrated for the first time, with efficiency ~10-14%  Multi-pass experiment – SPS ring More complex problem with not so clear outcome lack of knowledge on physics of nuclear interactions involved in multipass extraction Narrower angular scan (suppressed contribution of multipass extraction?) Lower deflection efficiencies (up to 10%) and bigger spread in values for different configurations  Open issues EMD suppression not proved experimentally (neutron loss?) Radiation damage to crystal not investigated(much lower limit expected than for p) Multi-pass ion interactions not clear (Si~amorphous material if channelling conditions are not satisfied.. )

16. 16/37 April 2007Walter Scandale At crystal Lambertson, crystal E853: extraction of 900 GeV protons (Tevatron: 1993-98)  Extracted significant beams from the Tevatron parasitic, kicked and RF stimulated  First ever luminosity-driven extraction  Highest energy channeling ever  Useful collimation studies  Extensive information on time-dependent behavior  Very robust

17. 17/37 April 2007Walter Scandale Crystal collimation at RHIC  Indirect experiment (measure particles disappearance) with Au and p runs  Si crystal 5×1 mm with  B =465 mrad located in interaction region matching section  Positioning not optimal (large beam divergence and  ≠ 0)  Crystal bends in the same plane where it scrapes  sensitivity to horiz. halo No clear interpretation of the results!  Measured ch. efficiency (~25%) doesn’t match theoretical predictions ( 56% with nominal machine optics). Better agreement and consistency when using measured beam divergence  need accurate knowledge of lattice functions.  Multipass physics and halo distribution models too simplistic?  Low channelling efficiency  collimation not successful & increased backgrounds !! R.Fliller III, A.Drees, 2005 4 crystal scans with different scraper positions - x s STAR Background during crystal collimation Not conclusive & abandoned !

18. 18/37 April 2007Walter Scandale Crystal Collimator in E0 replacing a Tungsten Target (2005) Crystal collimation at FNAL Using the crystal, the secondary collimator E03 can remain further (-1 mm or so) from the beam and achieve almost a factor of 2 better result! Tungsten scatterer Crystal

19. 19/37 April 2007Walter Scandale goniometer in 2006 goniometer in 2007 bending magnets upstream Si det 4 m The H8 line

20. 20/37 April 2007Walter Scandale S1 Si microstrips (AGILE) S3 GC S5 vacuum Si microstrips (AMS) p S2 70 m H S4 S6 B5 B6 Goniometer with crystal holders B 6 vacuum Q 14 vacuum 50 m 3-4 m vacuum p S2 The H8-RD22 apparatus 2006 Variant for 2007 Angular resolution

21. 21/37 April 2007Walter Scandale Strip silicon crystals  Strip Crystals have been fabricated in the Sensors and Semiconductor Laboratory (U. of Ferrara)  Mechanical bending exploits anticlastic forces Crystals sizes: 0.9 x 70 x 3 mm 3 and 0.5 x 70 x 3 mm 3 beam Main bending Anticlastic bending

22. 22/37 April 2007Walter Scandale Quasi-mosaic silicon crystals Crystal plate sizes: ~ 1 x 30 x 55 mm 3 critical angle for 400 GeV/c protons:  c ≈ 10  rad O.I.Sumbaev (1957) Quasi-Mosaic Crystals fabricated in PNPI (Gatchina, Russia) the mechanical bending of the crystal induces the bending of the atomic planes (initially flat and normal to large faces of plate) due to anisotropy  depends on the choice of crystallographic plane and on the angle of n 111 respect to the crystal face Quasi-mosaic bending Anticlastic bending Main bending R

23. 23/37 April 2007Walter Scandale Silicon detector Goniometer Granite Block Crystals Scintillator High precision goniometer

24. 24/37 April 2007Walter Scandale AMS Silicon Detectors Detector upstream of the crystal (on the granite block):  1 double-sided silicon microstrip detector: Resolution ~ 10  m in bending direction (X coordinate) Resolution ~ 30  m in non-bending direction (Y coordinate) Active area ~ 7.0 x 2.8 cm 2 Detector downstream of the crystal (on the granite block) :  1 BABY double-sided microstrip detectors (IRST): Resolution better than 10  m in bending direction Resolution better than 20  m in non-bending direction Active area ~ 1.9 x 1.9 cm 2 DOWNSTREAM TELESCOPE (at 65 m from crystal location):  4 AMS LADDERS: Resolution ~ 10  m in bending direction Resolution ~ 30  m in non-bending direction Active area ~ 4 x 7 cm 2 Silicon thickness: 300  m

25. 25/37 April 2007Walter Scandale AGILE Silicon Detectors  Single-sided silicon strip detectors  Built by Agile (INFN/TC-01/006)  active area 9.5 x 9.5 cm 2  Spatial resolution: ~ 40  m at normal incidence (~ 30  m for tracks at 11°)  Silicon thickness: 410  m  Upstream detector (before goniometer) 2 silicon detectors at 90° (corresponds to 1 X-Y plane)  Downstream detector 1 (at 65 m from crystal location): 4 X-Y silicon planes  Downstream detector 2 (at 65 m from crystal location): 6 X-Y silicon planes interleaved with 300  m tungsten planes

26. 26/37 April 2007Walter Scandale Gas Chamber and Scintillators  Gas Chamber Parallel plate chamber 0.6  12.8 mm 2 active area filled with Ar 70% + CO 2 30% 64 strips (pitch equal to 200  m) mounted on X-Y table able to withstand rates up to 10 8 ppp  Scintillating detectors Finger scintillators: 0.1  1  10 mm 3 Scintillating hodoscope: 16 strips with 2  4  30 mm 3 read-out by MAPMT (fast beam monitoring) Scintillator plates 100  100  4 mm 3 used for triggering silicon detectors

27. 27/37 April 2007Walter Scandale Angular scan of a crystal (1) Predictions in 1985-’87 by A.M.Taratin and S.A.Vorobiev, and O.I. Sumbaev Reflected Channeled d U Theoretical explanation of channeling and volume reflection phenomena Involved processes:  channeling  volume capture  de-channeling  volume reflection

28. 28/37 April 2007Walter Scandale Angular scan of a crystal (2) Results of the angular scan with Strip Crystal measured volume reflection angle: ~ 10  rad

29. 29/37 April 2007Walter Scandale measured volume reflection angle: ~ 10  rad Angular scan of a crystal (3)

30. 30/37 April 2007Walter Scandale Scan of Quasi-Mosaic Crystal  Orientation (111)  Bending angle: ~ 80  rad  Crystal sizes: 30 x 58 x 0.84 mm 3 measured volume reflection angle: ~ 10  rad

31. 31/37 April 2007Walter Scandale Double Reflection on Quasi-Mosaic Crystals (1) Experimental setup:  exploited rotational stage for off-axis alignment of the first crystal (preliminary scan)  used upper linear stage for alignment of second crystal  many steps for finding perfect alignment conditions

32. 32/37 April 2007Walter Scandale Double Reflection on Quasi-Mosaic Crystals (2) double reflection angle: ~ 20  rad

33. 33/37 April 2007Walter Scandale Double Reflection on Quasi-Mosaic Crystals (3) Chan1 unperturbed Chan2 Refl1 Refl2 Misaligned crystals -> two reflections angle: ~ 10  rad

34. 34/37 April 2007Walter Scandale Conclusive remarks  First observation of Volume Reflection Effect in bent silicon crystals with 400 GeV/c protons with efficiency close to unity  Measurement of volume reflection angle: ~ 10  rad  First observation of Double Reflection using two crystals in series: combined reflection angle is ~ 20  rad and efficiency close to 1  Channeling and Volume Reflection phenomena studied with Strip and Quasi-Mosaic Silicon Crystals (different fabrication techniques)  Measurement of crystals with different crystalline planes orientations: (111) and (110) Many thanks for the financial support to CARE-HHH INTAS-CERN INFN-CSN1 and INFN-NTA CERN AB & AT departments Many thanks to Efthymiopoulos, L. Gatignon, C. Rembser, R. Maccaferri SPS-OP team

35. 35/37 April 2007Walter Scandale  Crystal efficiency for channeling and volume reflection at both LHC injection (450 GeV) and top (7 TeV) energies.  Probability spectrum of proton deflections at 0.45 to 7 TeV for all physics processes down to a 10 -5 level.  Damage limit of crystals for instantaneous shock beam impact at ~15 MJ/mm 2.  Damage limit of crystals for integrated dose at ~5×10 16 p/year at 7 TeV.  Handling of crystals during normal operation at high-power impact.  Requirements for alignment and operational set-up (goniometer conceptual design, tolerances, time, etc.).  ……. Ion specific concerns: Si/ion interactions fairly understood for single pass channelling experiments.. what about multi-pass case? Crystal behaves like amorphous material for small impact parameters (surface effects)  what happens to isotopes produced by nuclear interactions with the wrong rigidity? What is the role played by nuclear/EM interactions in the case of volume reflection? Open issues

36. 36/37 April 2007Walter Scandale Future plans 0.1  1  10 mm 3  CARE-HHH workshop CC 07 22-23 March 2007  Proposal of LARP to create a working group aiming at a crystal experiment in the Tevatron  7 weeks beam time requested in 2007 at the SPS protons (H8 beam-line) at 400 GeV electrons and/or positrons at 300 GeV ions during dedicated MDs  Investigate edge-effect  Test of multi-strip crystals (Ferrara Sensor and Semiconductor Laboratory) to increase the angle of volume reflection  Test of germanium strip crystals and possibly zeolites  Upgrade of goniometric system with cradle for investigation of axial channeling  Upgrade and refurbishment of existing silicon microstrip detectors in order to increase spatial resolution

37. 37/37 April 2007Walter Scandale ALICE contribution  ALICE is supporting the effort of H8-RD22 in 2007, especially for the ion run  Crucial help expected in detecting fragmentation and EM dissociation down stream of the crystals in the next October run with ions