1. Crystal Collimation at SSC and Tevatron CARE Workshop on Crystal Collimation in Hadron Storage Rings CERN March 7-8, 2005 Fermilab CARE CC-2005 Nikolai Mokhov, Fermilab

2. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov2 OUTLINE Introduction Beam Collimation at Hadron Colliders First Look at Crystal Collimation for SSC Full-scale Modeling for Tevatron Proposal for Crystal Collimation at Tevatron Implementation and Commissioning at Tevatron

3. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov3 INRODUCTION Bent-crystal technique is well established for extracting high energy beams from accelerators. It was successfully applied at up to 900 GeV in E-853 at Fermilab, and simulations were able to predict the results correctly. Experiments at IHEP Protvino have demonstrated that 50-70% of the beam can be extracted using a thin (3-5 mm) Si channeling crystal with bending of 0.5-1.5 mrad. It was shown (SSC, 1991) that it is promising to apply this technique to a beam halo scraping at high energy colliders. A bent crystal, serving as a primary element, should coherently bend halo particles onto a secondary collimator. Based on realistic modeling (1999, 2003), it was proposed to implement a bent crystal into the Tevatron collimation system. It was done, and commissioning is planned in March 2005.

4. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov4 BEAM COLLIMATION AT COLLIDERS Beam collimation is mandatory at any superconducting hadron collider to sustain favorable background conditions in experiments, maintain operational reliability in stores and over the life of the machine, protect components against excessive irradiation and possible catastrophic damage, and reduce the impact of radiation on the environment. Only with a very efficient beam collimation system can one reduce uncontrolled beam losses in the machine and detectors to allowable levels. Bent crystal provides a possibility to substantially improve collimation efficiency.

5. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov5 SSC BEAM SCRAPER SYSTEM (1991) Primary H/V scrapers with/without target followed by a set of secondary collimators at appropriate phase advances

6. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov6 SSC SCRAPERS: 3 CASES STUDIED FOR 2 AND 20 TeV No target; copper (1.2-2 m), tungsten (0.8-1m), and graphite (4m) scrapers. 0.5-1 mm tungsten target at the front of copper scraper. 5-mm silicon bent crystal at the front of copper scraper. Multi-turn STRUCT/MARS calculations of beam loss distributions and heat loads to SSC components.

7. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov7 IMPACT PARAMETER ON SCRAPER AT 20 TEV 1.No target: 1-2  m. 2.With 1 mm tungsten target: 100-150  m. 3.With 5-mm silicon bent crystal: up to 500  m.

8. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov8 COLLIMATION INEFFICIENCY AT SSC CaseElastic  p/p  0.001  p/p  0.01  p/p  0.3 No target4.4600.0260.0532.520 1-mm W0.2400.0200.0310.175 5-mm bent Si 0.005000.007 Fraction of outscattered protons (%) per one 20 TeV proton: Calculated beam losses in SSC lattice were drastically down with W target and especially with bent crystal.

9. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov9 ENERGY DEPOSITION IN CU SCRAPER AT 20 TEV

10. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov10 ABSORPTION EFFICIENCY OF COPPER SCRAPER AT 20 TEV

11. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov11 TEVATRON COLLIMATION SYSTEM EVOLUTION Design report, commissioning, initial operation: a few single 0.9 to 1.8-m long SS collimators in front of SC magnets (Edwards, Pruss, Van Ginneken, 1979-1984). A set of two-unit collimators at optimal locations based on STRUCT/MARS modeling: 5-fold increase of 800-GeV proton beam intensity at fast resonant extraction (Drozhdin, Harrison, Mokhov, 1985). First two-stage system, two 2.5-mm thick L-shape tungsten targets with 0.3-mm offset relative to A0 scrapers: 5-fold reduction of beam loss rates upstream D0 and CDF detectors (Drozhdin, Mokhov et al., 1995). Genuine two-stage system proposed for Run-II with primary and secondary collimators at optimal locations optimized in STRUCT/MARS runs (Church, Drozhdin, Mokhov, 1999). Current system with tertiary collimators (Drozhdin, Mokhov,Still). Crystal primary collimator to be commissioned in March 2005.

12. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov12 TEVATRON RUN-II COLLIMATION Currently for proton beam, we have additionally D17(3) H/V, A11 (V) and A48 (V). Last two are for abort kicker prefire.

13. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov13 PROTON/ANTIPROTON BEAMS ON TEVATRON COLLIMATORS

14. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov14 BENT CRYSTAL FOR TEVATRON COLLIMATION (‘99) Biryukov, Drozhdin, Mokhov (PAC99) have shown how a silicon bent crystal can improve the Tevatron collimation system efficiency. Two cases were compared for a 900 GeV proton beam: the Run II collimation system with only one of three primary collimators—(D17h) horizontal—used. It intercepts large- amplitude protons and protons with positive  p; the same collimation scheme, but a silicon bent crystal is used instead of D17h. In reality, two additional primary collimators (bent crystals) should be used at D17v and D49h locations for vertical amplitude and off- momentum scraping. Therefore, results presented here represent about 30% of total losses in the machine.

15. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov15 CDF AND DØ DETECTORS AND ROMAN POTS Our studies have shown that the accelerator related background in the DØ and CDF collider detectors is originated from beam halo loss predominantly in the inner triplet region. In addition to the optically small aperture at  max location, the aperture restrictions in this area are the DØ forward detector’s Roman pots placed at 8  and the BØ Roman pots placed at 10  at the entrance and exit of the beam separators.

16. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov16 USING BENT CRYSTAL IN TEVATRON A silicon (110) crystal bent at an angle of 0.1- 0.3 mrad is placed upstream of the D17 secondary collimator instead of the original thin scattering tungsten target in the same position. Crystal channeling is simulated with the CATCH code. Particle tracking in the lattice is done with the STRUCT code with updated MARS physics. A non-channeling amorphous layer on the crystal surface due to its irregularity at a micron level is taken into account as a silicon target upstream of the crystal.

17. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov17 BENT CRYSTAL MODEL (1999)

18. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov18 BEAM LOSS REDUCTION WITH CRYSTAL 10-fold hit rate (beam loss) reduction in DØ and BØ. 4-fold reduction of radiation loads on downstream SC magnets.

19. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov19 PROTON DISTRIBUTIONS ON CRYSTAL

20. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov20 CRYSTAL ALIGNEMENT The crystal critical angle is ±5 µrad, therefore the efficiency depends strongly on the crystal alignment. With the alignment of -(104 - 111) µrad the large amplitude protons are captured by the crystal over the next 32 turns after the first scattering. For poorer alignment it takes longer time for the scattered protons to get into the critical region, which increases background in the detectors.

21. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov21 CRYSTAL LENGTH Angular distribution of protons after scattering on the amorphous layer depends on the crystal length. Shorter crystal would give smaller particle divergence, which should improve the system efficiency. In reality, a combined effect of scattering, channeling and tracking in the lattice could smear such a simple dependence. A shorter crystal indeed is better for the CDF Roman pots if its length <5 mm, but for longer crystals and for the DØ detectors the results obtained are almost independent of length.

22. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov22 BEAM LOSS DISTRIBUTIONS Overall beam loss distributions along the Tevatron lattice are shown on next page for the two studied cases with a 5 mm thick tungsten target and a 5 mm thick silicon bent crystal as the D17 primary scatterer at 5 . The secondary collimators are installed at 6 . Beam loss on the primary element itself is not shown. One sees that not only beam loss is lower at the collider detectors at BØ and DØ, but the entire machine becomes cleaner if the bent crystal is used.

23. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov23 BEAM LOSS with W (top) and Si (bottom)

24. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov24 CONCLUSIONS FROM 1999 STUDY The studies performed have shown that a replacement of the amorphous tungsten target as a primary collimator in the Tevatron beam collimation system with a 5 mm thick silicon bent crystal would reduce by about one order of magnitude the accelerator-related backgrounds in the CDF and DØ detectors and decrease beam losses in the supercondcting magnets of the D sector by a factor of four.

25. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov25 MOTIVATION FOR CRYSTAL COLLIMATION Goal CDF halo loss limit Growing concern that with the increase of beam intensities, proton losses due to beam-beam effects during a store, would hinder detectors from taking data for a portion of a store Tevatron: Further increase of collimation efficiency is needed ! Courtesy D. Still

26. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov26 Courtesy D. Still

27. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov27 PROPOSAL FOR CRYSTAL COLLIMATION (2003) Current collimation system in Tevatron is somewhat different compared to the one planned before Run-II. Based on detailed modeling, Carrigan, Drozhdin, Mokhov and Still, proposed to implement a bent crystal in the EØ straight section.

28. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov28 BENT CRYSTAL IN EØ Current primary H-collimator (D49 tungsten L-shaped target) is before the dog-leg at  H =96 m, D=2.3 m. The crystal is in the dog-leg at  H =73 m, D=2.5 m, about the same phase advance wrt secondary collimators.

29. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov29 CHANNELED AND SCATTERED PROTONS ON EØ SECONDARY COLLIMATOR

30. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov30 BEAM LOSS: CRYSTAL vs TARGET Substantial improvement for DØ and modest effect for BØ (CDF)

31. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov31 O-shaped 110 Si-crystal 5-mm long, 5 mm H, 1mm V bending angle 0.439 mrad miscut angle 0.465 mrad Courtesy D. Still

32. CARE CC-2005 – CERN, March 7-8, 2005Crystal Collimation - N.V. Mokhov32 SUMMARY Detailed simulations show that it is promising to use a bent crystal as a primary collimator (target) to increase efficiency of beam collimation at hadron colliders. With the proposed single crystal implementation at Tevatron, a scraping efficiency is expected to be increased by about an order of magnitude, and machine- related backgrounds in CDF and DØ be reduced by at least a factor of two (a second crystal is needed and beam gas scattering upstream the detectors is unaffected by crystal). Commissioning at 150 and 980 GeV is planned in March 2005. With optimized crystal-based collimation system at LHC, one can expect substantial reduction of beam loss rates and accelerator-related backgrounds in detectors.