aa r X i v : . [ phy s i c s . acc - ph ] A ug Photon Collider Technology Overview
V. I. Telnov
Institute of Nuclear Physics, Novosibirsk 630090, Russia
DOI: w ill be assigned In this conference paper, I review the present status and technical problems of the Photoncollider, as well as various additional applications of Compton scattering.
In this report, I provide an overview of the technical (and some political) aspects of the PhotonLinear Collider (PLC). The physics program at the PLC is discussed in K. M¨onig’s talk at thisconference [1].The photon collider based on the conversion of electrons at a linear collider to high-energyphotons through Compton scattering of laser photons has been discussed and developed sinceearly 1980s [2, 3]. A photon collider would be a very natural and relatively cheap supplement toa high-energy e + e − linear collider. It would allow the study of New Physics in two additionaltypes of collisions, γγ and γe with energies and luminosities close to those in e + e − collisions.A comprehensive description of the PLC is given in the TESLA TDR [4]; practically everythingregarding the photon collider at TESLA is valid for the PLC at the ILC. Further progress on thePLC after 2001 has been summarized in my talks at PHOTON2005 [5, 6] and PHOTON2007 [7].What’s new since 2007? Unfortunately, the future of the ILC is still highly uncertain. Alot depends on the physics results from the LHC, but even the discovery of a new physics atthe LHC would not guarantee the approval of the ILC (or CLIC?) construction due to its highcost (“high” as perceived by politicians). A possible way to overcome this barrier could be tobuild the linear collider in several stages. Recently, in October 2008, Prof. Hirotaka Sugawarasuggested to the ILC Steering Committee the construction of a “Photon collider Higgs factoryas a precursor to ILC,” as the required energy for producing a 120 GeV Higgs is lower in γγ collisions than in e + e − , positrons are not needed, and therefore such a collider would be muchcheaper. While laudable as an attempt to find a way out of the ILC stalemate, this suggestionhas caused concern to many in the ILC community because it would have meant an additionaldelay of 5-6 years in the start of e + e − operations at the ILC. After consultations with PLCexperts and additional study of the technical aspects and physics program for a low-energystartup scenario, the ILCSC rightfully concluded that it would more preferable to start with e + e − at 2 E = 230 GeV and investigate the Higgs in the e + e − → ZH process. This option’scost is not much higher than that of a 120 GeV PLC, but the physics case is stronger.The other interesting activities in the last few years related to the photon collider and basedon ideas originally proposed in the context of photon colliders are the developments of lasersystems for various applications based on Compton scattering.The outline of this paper is as follows. Basic properties of the PLC are considered in Sect. 2. PHOTON09
Let us consider briefly the main characteristics of backward Compton scattering important forthe photon collider.Kinematics. In the conversion region, a laser photon of energy ω collides with a high-energyelectron of energy E at a small collision angle α (almost head-on). The energy of the scatteredphoton ω depends on the photon scattering angle ϑ with respect to the initial direction of theelectron as follows [3]: ω = ω m ϑ/ϑ ) , ω m = xx + 1 E ; ϑ = mc E √ x + 1; x = 4 Eω m c cos α ≃ (cid:20) E TeV (cid:21) h µ m λ i , (1)where ω m is the maximum energy of scattered photons. For example: E = 250 GeV, ω = 1 . λ = 1 . µ m) (for the most powerful solid-state lasers) ⇒ x = 4 . ω m /E = 0 . γe collisions, the resulting luminosity spectrum can, in principle, be verynarrow (the electron beam collides with the most high-energy photons), while in γγ collisionsthe resulting luminosity spectra have the width at half-maximum of about 10-15% [8, 9].Maximum energy of scattered photons. With increasing x , the energy of the backscatteredphotons increases and the energy spectrum becomes narrower. However, at large values of x photons may be lost due to creation of e + e − pairs in collisions with laser photons, which leads toa reduction of the γγ luminosity [3, 8, 9]. The threshold of this reaction is x = 2(1+ √ ≈ . λ = 4 . E [TeV] µ m . Hence, the maximumenergy of photons at the PLC is about 0 . E .Polarization If laser photons are 100% circularly polarized, the backscattered photons atthe highest photon energy also have 100% circular polarization (even for unpolarized electronsand for any value of x ). The energy spectrum of scattered photons depends on the averageelectron helicity λ e and that of the laser photons P c . The relative number of hard photonsincreases when one uses beams with a negative value of the product λ e P c . For large x , thepolarization of electrons increases the number of photons in the high-energy peak almost bya factor of 2 (or 4 in the γγ luminosity). The energy spectrum of the scattered photons forvarious helicities of the electron and laser beams can be found elsewhere [3, 4]. A high degree ofphoton’s circular polarization is essential for the study of many physics processes, for example,for suppression of QED background in the study of the Higgs boson [4]. The ratio L /L (0,2is the total helicity of colliding photons) is larger when electron beams have a higher degree oflongitudinal polarization [4]. Modern electron guns give polarization up to ∼ PHOTON09 onlinear effects in the conversion. In order to convert nearly all electrons to high-energyphotons, the density of laser photons at the conversion point should be so high that the electroncan interact with several laser photons simultaneously. This nonlinear effect is characterizedby the parameter ξ = 2 n γ r e λ/α , where n γ is the density of laser photons, r e = e /mc and α = e / ~ c ([4] and references therein). The transverse motion of the electron in theelectromagnetic wave leads to an effective increase of the electron mass: m → m (1 + ξ ),which decreases the maximum energy of the scattered photons: ω m /E = x/ (1 + x + ξ ). TheCompton spectrum is shifted towards lower energies, higher harmonics appear, and the γγ luminosity spectra become broader. At x = 4 .
8, the value of ω m /E decreases by about 5% for ξ = 0 .
3, which can be considered the limit.Laser flash energy. For small conversion coefficients k = N γ /N e ∼ − exp( − A/A ), wherethe flash energy A is determined by the diffractive divergence of the laser beam and ge-ometric size of the electron beam. For head-on collisions and very narrow electron beams A ∼ ~ cσ z /σ c [9], where σ z is the r.m.s. length of the electron beam and σ c is the Comptoncross section. For x = 4 . A ∼ σ z [mm] J, while for x ≪ ⇒
1. For the ILC ( σ z = 0 .
3) mm this estimategives A ∼ k ∼ ξ small one should make the conversion length longer, which increases the required flashenergy. In calculating the flash energy, one should also take into account the collision anglebetween the laser and electron beams (when the laser optics is outside of the electron beam),the effective transverse size of the electron beam due to the tilt in the crab-crossing schemeof collisions and the angular size of the first quad (if optical mirrors are situated outside thedetector). A realistic calculation for ILC(500) gives A ∼ J [12, 11, 6]. . γγ , γe luminosities In e + e − collisions, the maximum achievable luminosity is determined by beamstrahlung andbeam instabilities. At photon colliders, the only effect that restricts the γγ luminosity is theconversion of the high-energy photons into e + e − pairs in the field of the opposing beam –that is, the coherent pair creation [13, 8]. For γe collisions, the luminosity is determined bybeamstahlung, coherent pair creation, and the beam displacement during the collision [9, 4].It is interesting to note that at the center-of-mass energies below 0.5–1 TeV and for electronbeams that are not too short (the case of ILC), coherent pair creation is suppressed due tothe broadening and displacement of the electron beams during the collision. For e + e − , theminimum horizontal beam size restricted by beamstrahlung is about 500 nm at the ILC, whilethe photon collider can work even with σ x ∼
10 nm at 2 E = 500 GeV, delivering a luminositymuch higher than that in e + e − collisions [14, 15, 4]. In fact, the γγ luminosity is simplyproportional to the geometric e − e − luminosity L geom .Unfortunately, the beam emittances in the damping-ring designs currently under consider-ation cannot achieve beam sizes that are smaller than σ x ∼
250 nm and σ y ∼ σ x by a factor of two seems possible. In principle, one can use electron beamsdirectly from low-emittance photo-guns, avoiding the need for damping rings altogether, butat present they offer a product of the transverse emittances that is noticeably larger than canbe obtained with damping rings: with polarized electron beams directly from photo-guns, theluminosity would be 100 times smaller! PHOTON09 L γγ ∼ . L geom [4].With “nominal” ILC beam parameters, the expected γγ luminosity in the high-energy peak ofthe luminosity spectrum L γγ ( z > . z m ) ∼ . × cm − s − ∼ . L e + e − [11, 6].Taking into account the fact that cross sections for many interesting processes are larger in γγ collisions than those in e + e − by an order of magnitude [4], the event rate in γγ collisionswith the nominal ILC beams would be similar, or perhaps somewhat larger, than in e + e − collisions. However, it is a highly unsatisfying situation to have the γγ luminosity limited bythe beam emittances, an order of magnitude below its physics limit determined by collisioneffects. It is an extremely interesting and important task to search for a realistic technicalsolution for obtaining beams with smaller emittances, and the first order of business shouldbe trying to optimize the damping rings for the specific requirements of achieving the highestpossible luminosity at the photon collider, as it was emphasized in [11, 6, 16]. Up to now theILC damping-ring design has been guided only by the baseline e + e − collisions.The typical γγ , γe luminosity spectra for the TESLA-ILC(500) parameters are shown inFig. 1 [4]. One can see that γγ and γe luminosities are comparable and these processes can bestudied simultaneously. However, it is much better to study γe collisions when only one of theelectron beams is converted to photons. In this case, one can measure the γe luminosity muchmore precisely [19]. The problem of measuring the γe luminosity spectra when both beams areconverted to photons is due to the uncertainty which direction the photon came from. For most dL 1dz L geom z = W / 2E γγγ e ILC(500) dL 1dz L geom z = W / 2E γγγ e ILC(500) Figure 1: γγ , γe luminosity spectra, left: both beams are converted to photons; right: onlyone beam is converted to photons.measurements, the luminosity as high as possible is desired. However, sometimes very clean γγ collisions with good monochromaticity and a reduced luminosity (in order to avoid overlappingevents) are needed. At large CP–IP distances and a non-zero crossing angle, the detector fieldserves as a deflecting magnet and allows more-or-less clean and quite monochromatic γγ , γe collisions to be obtained with a reduced luminosity, which will be useful to QCD studies, [20, 21].Luminosity stabilization. Beam collisions (luminosity) at linear colliders can be adjusted bya feedback system that measures the beam-beam deflection using beam position monitors andcorrects beam positions by fast kickers. This method is considered for e + e − collisions and isassumed for γγ as well [4, 6], though there are some differences between the e + e − and γγ cases.This problem and a stabilization algorithm were considered in detail in Ref. [6].4 PHOTON09 uminosity measurement. The measurement of the luminosity at the photon collider is notan easy task. The spectra are broad, and one should measure the luminosity and polarizationas a function of energies E , E of the colliding particles [19]. The luminosity spectrum andpolarization can be measured using various QED processes. These are γγ → l + l − ( l = e, µ ) [4,19], γγ → l + l − γ [19, 22] for γγ collisions and γe → γe and γe → e − e + e − for γe collisions [19].Some other SM processes could be useful as well.Absolute beam energy measurement. At the photon collider, the edge energy of the photonspectra and the electron beam energy E are not strictly connected due to nonlinear effectsin Compton scattering. The absolute energy calibration of the detector can be done using theprocess γe → eZ (during normal runs in γe mode or mixed γγ and γe mode) [23]. The general scheme of the photon collider is shownin Fig. 2. The optimum b ∼ γσ y , which is ∼ . σ y = 3 nm and 2 E = 500 GeV. This spaceis too small to fit any kind of a magnet for deflec-tion of used electron beams. In this case, there is amixture of γγ , γe and e − e − collisions. After cross-ing the conversion region, the electrons have a verybroad energy spectrum, E = (0.02–1) E and largedisruption angles due to deflection of low-energyelectrons in the field of the opposing beam. Theremoval of such a beam from the detector is there-fore far from trivial. quad crab crossing ~25 mrad~ EE E ~ (0.02−1) E IP. γ b l a s e r electronbunch C (e) (e)c . γ e e e αα c α γ Figure 2: Scheme of γγ , γe collider.The “crab crossing” scheme of beam collisions solves the problem of beam removal at photoncolliders [8, 9], Fig. 2 (bottom). In the crab-crossing scheme [25], the beams are collided at acrossing angle α c . In order to preserve the luminosity, the beams are tilted by a special RFcavity by the angle α c /
2. If the crossing angle is larger than the disruption angles, the beamsjust travel straight outside the quadrupoles.The disrupted beams after the IP have an angular spread of about ±
12 mrad [4, 11, 6].The disruption angle for low-energy particles is proportional to p N/σ z E [8, 9] and dependsvery weakly on the transverse beam sizes. The required crossing angle is determined by thedisruption angle, the outer radius of the final quadrupole (about 5 cm [11, 6]), and the distancebetween the first quad and the IP (about 4 m), which gives α c = 12 + 5 / ≈
25 mrad.In the present ILC design [10] only one IP is planned, with a crossing angle of 14 mrad andtwo detectors in the pull-push configuration. On the other hand, at the photon collider thecrossing angle should be at least 25 mrad. At first sight, it would therefore seem quite reasonableto design the ILC with 25 mrad crossing angle both for the e + e − and the photon collider.However, it was decided to make different collision angles due to very different requirements tothe extraction lines and beam dumps. In the e + e − case, after collision the beams remain quitemonochromatic, and so there is a possibility to measure their properties (the energy spectrumand polarization). At the photon collider, the situation is different: 1) the disrupted beams ata photon collider consist of an equal mixture of electrons and photons; 2) beams have a largeangular spread and need exit pipes of a large diameter; 3) the photon beam after the Comptonscattering is very narrow, it cannot be dumped directly at any solid or liquid material. There PHOTON09 e + e − and γγ . For the transition from e + e − to γγ , one has to move the detector andabout 700 m of the up-stream beamline [26]. The photon collider at ILC(500) requires a laser system with the following parameters [11, 6]:flash energy A ∼
10 J, σ t ∼ . λ ∼ µ m( ≤ E [TeV] µ m in a general case), and the ILCpulse structure: 3000 bunches within a 1 ms train and 5 Hz repetition rate for the trains, thetotal collision rate being 15 kHz.In addition to the average repetition rate, the time structure is of great importance. Theaverage power required of each of the two lasers for the photon collider at the ILC is 10 J × ∼
150 kW; however, the power within the 1 msec train is 10 J × / . ∼ O (1$)/W, the pumping efficiency about 25%, so the cost ofjust the pumping diodes would be O ($100M).Fortunately, at the PLC and other Compton scattering applications the same laser bunchcan be used multiple times. The most attractive approach is a stacking optical cavity that ispumped by a laser via a semi-transparent mirror [14, 17, 15, 18, 4, 12]. The ILC pulse structure(3000 bunches in the train with inter-pulse distance 100 m) is sufficient to create a ring cavityaround the detector. One can create inside such a cavity a light pulse with an intensity that isby a factor of Q (the quality factor of the cavity) greater than the incoming laser power. Thevalue of Q achievable at such powers is several hundreds and (even Q > Q .The external optical cavity (pulse stacking cavity) idea has proven to be a highly usefultechnique for HEP and other application (see the last Section). Recently at LAL, F. Zomer’sgroup has received in a Fabry-Perot cavity an enhancement factor of 10000! [27] They foundthat a simple concentric Fabry-Perot cavity is very unstable and sensitive to displacements.Much more stable is the 2D (planar) concave 4-mirror system, but it also has a problem:astigmatic and only linearly (or elliptically) polarized eigenmodes due to different reflection forthe s and p waves. A possible solation: 3D 4-mirror cavity that has reduced astigmatism andstable circularly polarized eigenmodes. The LAL group working in collaboration with Japanesecolleagues (see T. Takahashi’s talk in these proceedings) is in the process of developing sucha 3D 4-mirror cavity, plans to install it at the KEK ATF2 facility and obtain 1 MW averagepower in the cavity. For the photon collider, very stable both circular and linear polarizationsare needed. We see that this is not a simple task.Recently, J. Gronberg and B. Stuart from LLNL have proposed a plan on possible stagesin the development and construction of the laser system for the PLC [28]. They demonstratethat all the necessary technologies already exist. Pulse injection and intermediate amplificationdevices are off-the-shelf technologies; the main amplifier is not commercially available but atLLNL all required technologies exist (the Mercury laser is an existence proof). Gronberg andStuart specified six stages, where the first one is the pre-conceptual design and the last one isthe construction of a full-scale cavity and demonstration of its operation. The rough estimate6 PHOTON09 f the cost of the laser system is $20 M (“once it is known technology”). These are very niceplans! (dreams?) Unfortunately, at present there are no resources for such a program, partiallydue to the very uncertain plans on the ILC.In summary: at present, practically all laser technologies and components required for aphoton collider are in existence; nevertheless, the construction of such a state-of-the-art lasersystem will not be an easy task. The next step will be the development (on paper) of a detailedlaser scheme, its optimization, analyses of tolerances, methods of stabilization, figuring out whatalready exists and is known and what should be experimentally verified. Current developmentof passive stacking cavities for various applications based on Compton scattering is very helpfulfor the PLC.
We have routinely assumed that a linear collider would start with e + e − , while γγ , γe collisionswould arrive several years later. However, on many occasions many people have suggested tobuild the photon collider before e + e − (or even without e + e − ) because it is simpler (no e + , maybe no damping rings) and a somewhat lower beam energy is needed to produce an intermediate-mass Higgs boson. Several such suggestions (not all) are 1) V. Balakin et al. (1993)[29], basedon the VLEPP and D. Asner et al. (2001) [30], based on CLIC-1 with E = 70 GeV.Recently, in October 2008, H. Sugawara gave a talk at the ILCSC meeting entitled “Photon-photon collider Higgs factory as a precursor to ILC”. His main motivation was the following:the ILC(500) is too expensive, so let us build first a collider for the smallest reasonable energy.If the physics goal is H(120), then it could be produced in γγ → H at 2 E ∼
160 GeV, while e + e − → ZH needs higher energy. Besides, γγ does not need positrons. Before Sugawara’spresentation at the ILCSC meeting, one of the ILCSC members sent me the slides of the pro-posal and asked for an opinion on this subject. My reply to the ILCSC and to Sugawara-sanwas the following (shortly): • the cost of such a PLC will be not much cheaper because it needs damping rings withsmall emittance and polarized electrons. Polarization is absolutely necessary for the Higgsstudy. A polarized electron gun with small emittance does not exist, therefore dampingrings are unavoidable; • the laser system for the PLC is not simple, its developments have not started yet; • the H(120) can be be studied much better in e + e − → ZH at 2 E = 230 GeV; • the ILC community will never agree with a proposal that shifts the start of e + e − experi-ments by about 5 years; • the PLC gives a unique possibility to study new physics at the LC in two additional modesat a small additional cost, gives access to higher masses, but it would be better to planthe PLC as the second stage of the ILC (as it was usually assumed).After the meeting, the ILCSC requested the LOI Physics pannel to look in more detail intothe physics case, and the GDE to look at machine designs for this kind of a staged approachto ILC construction and operation. In January 2009, T. Barklow, J. Gronberg, M. Peskin andA. Seryi (BGPS) prepared a draft of the report, which was discussed at the expanded Physicspanel with invited PLC experts and was not supported.In February 2009, the BGPS report was reviewed at an ILCSC meeting and the conclusionwas the following: “A 180 GeV gamma-gamma precursor would cost about half that of the PHOTON09
700 GeV ILC, but would produce much less physics. A better alternative for early Higgsstudies would be a ∼
230 GeV e + e − collider for studying the Higgs through ZH production;this would be about 30% more costly than the γγ collider. ILCSC decided not to pursue thegamma-gamma collider (as the ILC precursor) further at this time.”The PLC as the first stage of ILC was also discussed at TILC09 in April 2009: T. Barklow’splenary talk [31] and the discussion at Joint Physics session with my introductory talk [32].All people have agreed that PLC is certainly necessary, but it would be better to start with e + e − (230). Since e + e − (230) is needed in any case, a PLC precursor results in no cost reductionat all. See also J. Gronberg’s talk at PHOTON-2009 [33]. This is all about the PLC as aprecursor to the ILC.Nevertheless, proposals to build some type of a ”cheap” low-energy PLC continue to resur-face. It could have made sense if the production rates for some particles were higher thanthose at e + e − factories. However, simple estimates show that e + e − B-factories (and especiallya future Super B factory) have much higher productivity for any particles of interest. So, thePLC should wait for its logical turn at the ILC (or CLIC, or other LC).
In 1981, when the PLC was proposed, the maximum e → γ conversion efficiency in Comptonscattering of a laser light was at the level k = N γ /N e ∼ − . For the PLC, k ∼ e + generation).Let us enumerate some of the applications: • Medical applications: – Dichromatic imaging: illuminate above and below contrast K-edge, digital imagesubtraction. Established at synchrotrons (access limited, expensive $100s); – Computer tomography with monochromatic X-rays. In mammography conventionalX-ray imaging is difficult, soft-tissue contrast is poor. Monochromatic X-rays enablenew techniques: phase contrast imaging 3D with low dose; • Fast X-ray materials characterization: composition of materials, inspection of trucks andcontainers; • Nuclear materials detection (in trucks and containers) by monochronatic γ beam; • Defect profiling with e + s. MeV photons produce positrons which gather at defects, detec-tion of two 510 keV annihilation photons shows the source position (PET-tomography).Directly probe material defects. • Nuclear waste assay. Resonance scattering of 1-5 MeV γ -quanta is a unique fingerprintof nuclides, radioactive and stable nuclides can be detected;8 PHOTON09
Beam diagnostics and polarimetry at electron accelerators; • Obtaining polarized e + for e + e − linear colliders [35]. Similar to the undulator source ofpolarized γ based on ∼
150 GeV main linac beams but needs much lower electron ener-gies and is independent of the main collider. There is a very active collaboration namedPOSIPOL, which has already conducted four workshops on this topic; • Laser cooling of electrons [36] can considerably reduce emittances of beams and increasethe luminosity of photon colliders, provide a fast beam cooling in damping rings [37] forX-ray production and for e + production [35]; • The Photon Collider.Already many ICS facilities are under construction in the world [38]. These activities (allenumerated above and some others) are called by people as the ”Compton World Wide Web ofLaser Compton.” Let us hope that the Photon Collider will be eventually constructed somewherein the world as well!
References [1] K. M¨onig, talk at Photon09, these proceedings.[2] I.F. Ginzburg, G.L. Kotkin, V.G. Serbo, and V.I. Telnov, Pizma ZhETF,
514 (1981);[JETP Lett. , 491 (1982)].[3] I.F. Ginzburg et.al., Nucl. Instrum. Meth. , 47 (1983); ibid A219 , 5 (1984).[4] B. Badelek et. al., Intern. Journ. Mod. Phys.
A30 , 5097 (2004), hep-ex/0108012.[5] V.I. Telnov, Acta Physica Polonica,
B37 , 633 (2006), physics/0602172.[6] V.I. Telnov, Acta Physica Polonica,
B37 , 1049 (2006); physics/0604108.[7] V.I.Telnov, Nuclear Physics B (Proc. Suppl.)
271 (2008).[8] V.I. Telnov, Nucl. Instrum. Meth.,
A294 , 72 (1990).[9] V.I. Telnov, Nucl. Instrum. Meth.
A355 , 3 (1995).[10] ILC Reference Design Report,ILC-Report-2007-001, arXiv:0712.2361 [physics.acc-ph].[11] V.I. Telnov, Proc. of 2005 Intern. Linear Collider Physics and Detector Workshop and2nd ILC Acceler. Workshop, Snowmass, Colorado, 14-27 Aug 2005, ECONF C0508141:PLEN0020,2005, physics/0512048.[12] G. Klemz, K. Monig and I. Will, Nucl. Instrum. Meth.
A564 , 212 (2006), physics/0507078.[13] P. Chen and V. I. Telnov, Phys. Rev. Lett., , 1796 (1989).[14] V.I. Telnov, Nucl. Phys. Proc. Suppl. , 359 (2000), hep-ex/9908005.[15] V.I. Telnov, Nucl. Instrum. Meth., A472 , 43 (2001), hep-ex/0010033.[16] V.I. Telnov, Pramana Journal of Physics, , 957 (2007), physics/0610285.[17] V.I. Telnov, Int. J. of Mod. Phys. A15 , 2577 (2000), hep-ex/0003024.[18] I. Will, T. Quast, H. Redlin and W. Sandner,
Nucl. Instrum. Meth.
A472 , 79 (2001).
PHOTON09 , 379 (2004), hep-ex/0301037.[20] V.I. Telnov, talk at the ECFA workshop on linear colliders, Montpellier, France, 12-16November 2003; slides: [21] V. I. Telnov, Nucl. Phys. Proc. Suppl. (2008) 81.[22] V. Makarenko, K. Monig and T. Shishkina, Eur. Phys. J.
C 32 , (2003) SUPPL1143.[23] V.I.Telnov, Calibration of energies at the photon collider, talk at TILC09, April 17-21,2009, Tsukuba, Japan. http://tilc09.kek.jp [24] L.I. Shekhtman and V.I. Telnov, Proc. of Intern. Conf. on Linear Colliders (LCWS 04),Paris, France, 19-24 Apr 2004,
Paris 2004, Linear colliders , v1, p.507; physics/0411253.[25] R. B. Palmer, In
DPF Summer Study Snowmass ’88: High Energy Physics in the 1990’s,Snowmass, Colo., Jun 27 - Jul 15, 1988 , SLAC-PUB 4707.[26] V.I. Telnov, Proc. LCWS06, India, March 2006. Pramana Journal of Physics, , 1177(2007), physics/0610287.[27] F. Zomer, A. Jeremie, EUROTeV-Report-2008-096-1.[28] J.Gronberg and B.Stuart, Photon Collider Laser Work at LLNL, talk at TILC09, April17-21, 2009, Tsukuba, Japan. http://tilc09.kek.jp [29] V. E. Balakin and I. F. Ginzburg, Proc. 2nd Intern. Workshop on Physics and Experimentswith Linear e + e − Colliders, Waikoloa, Hawaii, 26-30 Apr, 1993, v2, p. 605.[30] D. Asner et al. , Eur. Phys. J. C (2003) 27, hep-ex/0111056.[31] T. Barklow, Gamma-gamma collider physics report, talk at TILC09, April 17-21, 2009,Tsukuba, Japan, http://tilc09.kek.jp [32] V.I.Telnov, Introduction to the discussion on Physics case of the PLC as the first stage ofILC, talk at TILC09, April 17-21, 2009, Tsukuba, Japan, http://tilc09.kek.jp [33] J. Gronberg, Costs versus benefit of an early photon collider project, these proceedings.[34] D. Strickland and G. Mourou, Opt. Commun. , , 219 (1985).[35] S. Araki et al. , Conceptual design of a polarised positron source based on laser Comptonscattering, arXiv:physics/0509016. F. Zimmermann et al. , CLIC polarized positron sourcebased on laser Compton scattering, CERN-CLIC-NOTE-674; M. Kuriki et al. , ILCpositron source based on laser Compton, AIP Conf. Proc. (2008) 92.[36] V.I. Telnov, Phys. Rev. Lett., , 4757 (1997), Erratum: Phys. Rev. Lett. , 2747 (1998);V.I. Telnov, Nucl. Instrum. Meth., A455 , 63 (2000), hep-ex/0001029.[37] Z. Huang and R. D. Ruth, Phys. Rev. Lett. , 976 (1998).[38] Workshop on Compton sources for X/gamma rays: Physics and Applications, 7-12 Sept,2008, Alghero, Italy, http://agenda.infn.it/conferenceTimeTable.py?confId=36710