Nuclear Instruments and Methods in Physics Research A 355 (1995) 3-18 NUCLEAR

INSTRllWiNTS
&METNoDs
IN PNVSICS

ELSEYIER
REsgEP
Principles of photon colliders
Valery Telnov *
Ins&ate of Nuclear Physics 630090, Novosibirsk, Russian Federation

Abstract
Future linear colliders offer unique opportunities to study yy, ye interactions. Using the laser backscattering method one
can obtain yy, ye colliding beams with energy and luminosity comparable to the electron-position luminosity or even
higher. In this review physical principles of photon colliders are described and various problems, concerning the accelerator,
laser, interaction region and luminosity are discussed. Some examples of physical processes are given.

1. Introduction in Ref. [3]. Polarization effects have been considered in

Ref. [6]. Collision effects restricting the luminosities,
schemes of the interaction region, requirements for accel-
Since 1970 two-photon physics has been actively stud-
erators, attainable luminosities and other aspects of obtain-
ied at e+e- storage rings in collisions of virtual photons.
ing yy, ye collisions have been considered in Refs. [7,8].
Their spectrum is dn N O.O35do/w, so that two-photon
Physical problems, which can be studied in yy, ye colli-
luminosity is much lower than that in e+e- collisions.
sions, were discussed in many papers (see Refs. [lo-141
Nevertheless, these experiments have given us a lot of
and references therein).
information on the nature of eIementary particle
Very rich physics, potentially higher than efe- colli-
Future linear colliders which are being developed now
sions luminosity, and simplification of the collider
(see Table 1) offer unique, much more rich than before,
(positrons are not required) are all attractive to physicists.
opportunities to study yy and ye interactions. Unlike the
The progress in the development of high power lasers
situation in storage rings, in linear colliders each bunch is
(both conventional and free-electron lasers) makes it now
used only once. This makes it possible to “convert”
possible to seriously consider photon colliders.
electron to high energy photons to obtain colliding y-y and
ye beams. This idea was proposed by the author and
colleagues in 1981 [l] and was further discussed and
developed in Refs. [2-8,10]. Among various methods of 2. Linear colliders
e + y conversion [7] the best one is Compton scattering of
laser light on high energy electrons. This method is well It is well known that due to the synchrotron radiation
known 193. After scattering, the photons have an energy problem in e+e- storage rings the energy region beyond
close to that of the initial electrons and follow their LEP II can only be explored by linear colliders. In linear
directions. The small bunch size in linear colliders makes colliders beams are used only once which opens the possi-
it possible to achieve a conversion coefficient k = NY/N, bility to obtain yy, ye colliding beams.
_ 1 at a moderate laser flash energy of a few joules. In y-y With an achieved accelerating gradient of about 100
collisions a luminosity higher than in e+e- collisions is MeV/m and a reasonable length of lo-20 km the next
possible due to the absence of some collision effects. linear colliders can cover the energy region up to l-2 TeV
Monochromaticity of collisions near AW,,/W,, _ 10% which is comparable with the constituent energy poten-
can be obtained. Photons may have various polarizations, tially available at LHC. Due to the complexity of the task,
which is very advantageous for experiments. the linear collider community is considering now as a first
The detailed considerations of the conversion, photon step a linear collider with a c.m. energy up to 500 GeV
spectra and monochromatization of collisions can be found with a yearly integrated luminosity of about 20 fb-‘.
Table 1 gives some parameters essential for out task of
linear colliders now under consideration [15].
Beam parameters in the table have been optimized for
e+e- collisions but may be used as a reference point for
* E-mail teInov@inp.nsk.su. photon colliders.

0168-9002/95/$09.50 0 1995 Elsevier Science B.V. All rights reserved

SSDIO168-9002(94)01173-7 I. INTRODUCTORY PAPERS
4 V. Teltwv / Nucl. Instr. and Meth. in Phys. Res. A 355 (1995) 3-18

---
3. Backward Compton scattering, main characterictics
of photon beams and y y , ye collisions

If laser light is scattered off an electron beam, the

photons after scattering have a high energy (w -E,) and
follow the initial electron direction with additional angular
spread - l/y. This method of e + y conversion has
obvious advantages in comparison with other methods
(bremsstrahlung on amorphous or crystal target, beam-
strahlung during collision of e+e- beams) because of Fig. 1. Kinematics of Compton scattering.

much better background conditions, the possibility of

monochromatization (- 10% in yy collisions and even
better in ye collisions) and a high degree of circular
polarization.

3.1. Kinematics

Below, formulae for Compton scattering are given for

the case of our interest [3].
In the conversion region a photon with energy we is
scattered on an electron with energy E, at a small collision
angle eye (Fig. 1). The energy of the scattered photon o
depends on its angle 6 relative to the motion of the
incident electron as follows:
00
Cd= (1)
1+ (4/190)2 ’
where

%I
= XE,;
x+1
Fig. 2. Electron and photon scattering angles vs. photon energy for
lYo= +T;
0
Y = 4.8.

4E, w. cos2cio/2
.X= =11.3[&][5$
m2c4 The photon and electron scattering angles are unique
where w,,, is the maximum photon energy. functions of the photon energy:
For example: E, = 300 GeV, o. = 1.17 eV (neodinium
glass laser) *x = 5.37 and w/E, = 0.84. As will be ex- 19y(y)=t90 T-1, &=aL
li Yl-y’
plained below, the value x = 4.8 is optimum for photon
colliders; therefore further examples will be given for this where y = o/E,. These functions for x = 4.8 are dis-
value of x. played in Fig. 2.

Table 1
Some parameters of 0.5 TeV linear colliders
.--_-
VI FPP MC JLc(X) CLIC TESLA
(Russia) w-At) (=K) (CERN (DESY)
L, 1O33 crn-‘~-~ 12 8 6.3 2.2-8.9 3.7 6.5
Rep. rate, Hz 300 180 150 1700 50 10
# bunch/train 1 90 90 l-4 125 800
Part./bunch (10”) 20 0.65 0.63 0.6 2.9 5.2
a, (nm) 2000 300 260 90 670 1000
uy (nm) 4 3 3 8 30 65
uz (mm) 0.75 0.1 0.07 0.17 0.5 1
At bunch (ns) - 1.4 1.4 0.33 16
 V. Teltwv / Nucl. Ins@. and Meth. in Phys. Res. A 355 (1995) 3-18 5

7 L

c
The energy spectrum of the scattered photons is defined 1
- 1 da
I
by the cross section 4 R

6 uc dy X=4.8 II
II
II

x
+2A,P,rx(l-2r)(2 -y)
1;
Y= +_=-.
0 x+1’

Y 2 -
r=------1.
x(1-y) - ’
, ‘& ._z__‘:’ . . . . -

a,=a
i
7
t?ZC
ez

1
’
= 2.5 X 1O-25 cm=,

The electron mean helicity is A,(( A, 1I $1, and that of the

0 0.1 0.2
--__ --___--

0.3 0.4 0.5 0.6

Y = w/E,
0.7
:I

0.8
I
0.9

laser photons is P,.

The total Compton cross section is Fig. 3. Energy spectra of scattered photons for different polariza-

a, = u,“P + 2A,P,a,, (4)

a”P=22 l-4 --$

c
x x
[( 1
1
ln(x+l)+i+i 3.2. Choice of a laser wavelength

By increasing the energy of the laser photons, the

-
maximum energy of the scattered photons also increases
2(X + 1)”
and monochromaticity of the spectrum improves. How-

u,=~l+! x
[(
x
)
In(.x+1)-5+-
2
1
x+1
ever, besides Compton scattering, other processes become
possible in the conversion region [3,7]. The most important
one is the process of e+e- pair creation in the collision of

1.
1 a laser photon with a high energy scattered photon.
The threshold of this reaction is o,oo > m2c4, i.e.
2(x+1)2
In the region of our interest x = l-10 the ratio ( ‘+Ja,(
< 0.2, i.e. the total cross section only depends slightly on
the polarization. However, the energy spectrum does es- dN/dy X=4.8
sentially depend on the value of 2&P,. At 2A,Pc = - 1 2P,h=-1
and x > 2 the relative number of hard photons nearly
10 \
doubles (Fig. 3), improving essentially the monochromatic- 2

ity of the photon beam.

The photon energy spectrum presented in Fig. 3 corre-
sponds to the case of a small conversion coefficient. In a
“thick” target each electron may undergo multiple Comp-
ton scatterings. The secondary photons are softer in gen- 1
eral and populate the low part of spectrum. In Fig. 4 are
plotted the photon spectra for various thicknesses of a laser
target expressed in units of the collision length for elec- 0 0.5 1 2 3
trons of the initial energy. In the same picture the ratio of WJ,- = 1 1.46 2.13 3.8 6.1
the total number of produced photons to the number of
photons produced by primary electrons is also given. Mul- 10’1.~..‘....‘,...‘....‘,...‘....’....’....,.1.,
tiple Compton scatterings lead to an increase of yy lumi- 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 C9
nosity at low and intermediate invariant masses (that is Y = w/E,
positive), but, on the other hand, they spoil the polarization Fig. 4. Normalized (to maximum) photon spectra for different
properties of scattered photons (see Section 3.3) and also numbers of interaction lengths in a laser target (these numbers
produce background in the detector (y-y + hadrons, see mark the solid curves); dashed curve is photon spectra without
Section 9). secondary scatterings.

1. INTRODUCTORY PAPERS
6 V. Telnov/Nucl. In&. and Meth. in Phys.Rex A 355 (1995) 3-18

x > 2(1 + fi) = 4.83. At x = 4.8 the wavelength or the

energy of the laser photons are the following
A = 4.2E,,[TeV] p.m; w. = 0.3/E,[TeV] eV. (5)
Above the threshold region (x - 8-20) the pair produc-
tion cross section exceeds the Compton one by a factor of
1.5-2 (see Fig. 5 [7], [8]). Due to this fact the maximum
conversion coefficient at large x is limited by 25-30%.
For these reasons it is preferable to work at x - 4.8.

3.3. Polarization of photons

If electrons or laser photons are longitudinally polar-

ized, the scattered high energy photons have circular polar-
ization too. The mean photon helicity is given by [6]:

A,(y)= 2h,xr[l+(l-y)(l-2r)*] 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 ( 9
[
y=dL

+P.(l-2r)(&+l-Y)] Fig. 6. Helicity of scattered photons vs. o/E, for various

polarizations of laser and electron beams.

/I -
1-Y
+ 1 -y - 4r(l -r)
One can also produce photons with a transverse polar-
ization [6], but it is less useful than longitudital.
Polarization characteristics of the luminosity are dis-
+2P,A,m(l-2r)(2-y) 9
1 cussed in the next session. The polarization is crucial for
some experiments [16,10].
where y and r were defined before. This is shown in Fig.
6 for various helicities of electron and laser beams. Note,
3.4. Spectral luminosity, monochromaticity
if polarization of laser photons PC = f 1, then A, = -PC
at y = ym. In the case of 2 P, A, = - 1 (the case with good
The spectrum of the scattered photons is very broad,
monochromaticity) all the photons in the high energy peak
but because of the energy-angle correlation in Compton
have a high degree like-sign polarization.
scattering it is possible to obtain rather narrow distribu-
tions of the spectral luminosities of y-y and ye collisions.
1 , I2 The spectral luminosity distributions depends on the vari-

0.8 -
2.4 dLr/dz X=4.8 ,”
k2L. 1 I
I I
0.6 -

1.6
0.4 -

0.2 -
0.8

,
ot..“.‘..““.“.“..‘~I
0 2.5 5 7.5 12.5 10 15 17.5
0 0.4
X
0
Fig. 5. (a) The ratio of cross sections for e+e- pair creation in
the collision of laser and high energy photons and for Compton
scattering aad (b) dependence of the maximum conversion coeffi-
cient on x assuming w = w,. Fig. 7. Spectral luminosily of yy collisions.
 K Tebwv /Nucl. Instr. and Meth. in Phys. Res. A 355 (1995) 3-18 7

able p = b/ya, where b is the distance between the then at x 5 4.8, nevertheless in some experiments the
conversion region and the interaction point (i.p.1, and ultimate monochromaticity will be very useful. For exam-
a = a, is the r.m.s. radius of the electron beam at the i.p. If ple, the search for narrow resonances at large nonresonant
p +Z 1, then at the i.p. photons of various energies are background (Higgs, for example) requires / Ldt a l/A
mixed in space and the distribution in the invariant mass of and for resonance study / Ldt a 1/A2.
the yy of ye system is broad. But, if p Z+ 1, then in ye According to Fig. Sa the monochromaticity for polar-
collisions electrons collide only with the highest energy ized (2P,A, = - 1) beams is by a factor 2 higher than for
photons, therefore the invariant mass spectrum of a ye the unpolarized case. The difference is considerably higher
collision is narrow. In yy collisions at p Z+ 1 the photons when we compare absolute peak luminosities (dL,,/dz
with higher energy collide at a smaller spot size and, at z=z,, ) at the same distance b between conversion
therefore, contribute more to the luminosity. As a result, and interaction regions (see Fig. 8b). The ratio of peak
the luminosity spectrum is much narrower than at p < 1. differential luminosities is 4.4 and 7.4 for x = 5 and
For round beams the distribution of luminosity in in- x = 20 respectively. The total luminosity/beam collision
variant mass of yy system Wyy has the form [3] at p B 1 is

-
dL,, = 2zk2L,,j-ym f(x, Y)f( x7 Z2/Y) (10)
dz Z2/Y,

,,I’
,,
(7)
where z = W,,/2E,, p = b/ya, f(x, y) is defined in Eq.
(3), 1,,(x) is the Bessel function for an imaginary argument
and L,, = N2f/2na2. In Fig. 7 the plots of spectral
luminosities are shown for unpolarized and polarized beams
(2P,h, = - 1 both beams) for p = 0 and 1. One can see
that at p = 1 the luminosity in the low mass region is
strongly suppressed and the full width at half of maximum
is about 10% for polarized and 20% for unpolarized
beams.
Figures for the dependence of the monochromaticity
and yy luminosity on p2 can be found in Ref. [7]. 0 2.5 5 7.5 10 12.5 15 17.5 20
Let us consider here the ultimate case: p Z+ 1. In this 70 ,
case the spectral yy luminosity is given by the simple %,/dz 8zmax b)
formula [3] 60 (kZNzy2/4,#) at ‘=‘,

dL,, =- k2N%’ i
.’
pxb2 Z’f”(% z). 50 - a’
dz p=(b/w.) >>I .’
The monochromaticity of collisions can be character- .’ ’ 2P&,
40 r .’
ized by the value of #‘_ 0
.
: ___-_ _,
l dL,, _/’
-z,, at z=z,,. (9) 30 -
A=L YY dz
The reverse value 1)= l/A is approximately equal to the
relative full width of the luminosity distribution at half
20 -
r
,,, /’

maximum. The value of A is plotted in Fig. 8a as a

function of x. One can see that at x = 5 the width of
spectrum l/A _ 5% for polarized beams and 10% for
unpolarized beams and it improves with increasing x. In
10

O
E-
02.5 7.5 10 12.5
15 17.5 20
the case of polarized beams at x = 20 the monochromatic- X
ity is already lS%! Although at x = 20 the maximum Fig. 8. Peak differential y-y luminosity at p* = (b/yu,)’ B 1 VS.
conversion coefficient is only about 20% (Section 3.2), x: (a) normalized peak luminosity, (b) absolute peak luminosity
and the luminosity is by one order of magnitude lower for a fixed distance b between interaction and conversion regions.

1. INTRODUCTORYPAPERS
8 V. Telnou/Nucl. Instr. and Meth. in Phys. Res. A 355 (1995) 3-18

where, with 15% accuracy, + A,, A,, a L,, while for main background NYy_ Bsa 1
- A,, A,, a L, [16,10].
F(n) = 0.72 - 0.16 ln x for unpolarized beams, We have seen that for 2 P, A, - - 1 the ratio L,/L, is
F(x) = 1.25 -0.16 In n for 2P,h= -1. large in the high energy peak. In this case one can work at
p B- 1 with a good monochromaticity. For the opposite
At x = 4.8 for these two cases F = 0.42 and 1 respec- cases 2P,A, - + 1, the ratio L,/L, is large in a wide
tively. range of invariant masses (see Fig. lo), so, at first sight, it
The last remark about the case p >> 1 is that one may is convenient for the purpose of searching for the Higgs
think that an increase of p leads inevitably to a decrease of [lo]. But in this case, in order to have a broad spectrum,
the L,, luminosity. It is correct when we increase the one has to work at p x 1. However, in this regime the yy
distance between the conversion and the interaction region, luminosity may be much lower than at p >> 1 (if b is
in this case L,, a l/p’. But one can do the same by fixed and one artificially increases the electron cross sec-
decreasing the electron beam radius, in this case L,, --$ tion). Another important note is that the luminosity in the
const. as p + @J. intermediate invariant region is greatly populated by colli-
The possible helicities of two colliding photons is 0 or sions of photons produced in the secondary Compton
2, so d L,,/dt can be decomposed in two parts: dL,/dt scatterings; this considerably deteriorates the polarization.
and dL,/dz [6,34]. These luminosities are plotted in Figs. At last, as was pointed out by Richard [17] at this Work-
9a-9c for x = 4.8, PC = 1 and 2 A, = - 1, - 0.5, 0 (equal shop, there is a very large background for Higgs due to the
for both beams). The ratio L,/L, for the region 0.75 I z process yg + bb. Gluons in a photon carry in general a
5 zulax vs A, is shown in Fig. 9d. For 100% electron small part of the photon energy, and only operation with
polarization Lo/L, - 60 (by changing the polarization of good monochromaticity, at z -z,,,~, can eliminate this
one laser and the corresponding electron beam to the background.
opposite, one can get the reverse case). This is of great For ye collisions the formulae for luminosities can be
importance for Higgs measurements, because N_,+ n a 1 found elsewhere [3]. Note only that with an increase of p

1.4 d&/dz X=4.8 b

2.4

1.2
2

1.6

1.2
0.6

0.8

0.4

0
0 0.25 0.5 0.75 0.25 0.5 0.75
z=W,/2E,

1.4 :dL,,/dz X=4.8

c>
Pa=1
: k*l
1.2 0 - dwdz 2h=O
2 - dLJdz
1

0.8

0.6

0.4

0.2

0
0 0.25 0.5 0.75
z=w,/zE,

Fig. 9. (a-c) Spectral luminosities L, and L, for various degrees of electron polarization A, in the case of a laser helicity PC= 1 and
A, < 0; (d) ratio L,/L, at z > 0.75 vs. degree of the electron polarization.
 V. Telnov/ Nucl. Instr. and Meth. in Phys. Res. A 355 (1995)
3-18 9

of Compton scattering for the electron moving through a

laser target along the axis
2Au, m dz 2Au,
p=2 n,u,dz-- p=- (14)
/ i
matI, --30 Z2 chl, .
I+-
&
1.25
The factor 2 in the first equality is due to the relative
motion of electrons and photons. The effective interaction
length is given by the intergal and equals rrTTpy.The
probability p = 1 (A = A,) at

0.5 hCl, A,c tic2

A,== and P=-==. (15)
0.25
C ‘7 C
To provide the high photon density at the laser focus
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 during the collision with the electron bunch of length l,,
z=w,/zE, the laser bunch should have the length 1, > 1,. The mini-
mum laser flash energy and laser power needed for obtain-
Fig. 10. Spectral luminosities L, and L, for the case PC= 1 and ing the conversion coefficient k - 65% (A =A, in Eq.
A>O.
(11)) at x = 4.8 (a, = 9 X 10-25cm2) are
hC1,

the monochromaticity of ye collisions improves up to Ao,min= g- = 8.4Z,[cm] J, (16)

monochromaticity of electron beams (for any values of x). c
In this section we have considered only kinematical 2

characteristics of the luminosities. The collision effects, Po.min= y =; = 2.5 x 10” w. (17)
Y C
which determine attainable luminosity, are considered in
Section 5. These minimum values for A, and PO have been obtained
for a uniform photon density distribution along electron
and laser beams and j3, < I, = 1,. For Gaussian beams
4. Conversion coefficient, requirement of lasers with 1,=I,=2uZ>>/3,,

4.1. Estimation of the conversion coefficient Ae - &A,,,,, and Ppe* = fiPe,min. (18)
It is remarkable that the value of A, is almost independent
The conversion coefficient depends on the energy of of the focal spot size until 2& < I, = l,, i.e. aY
the laser flash A as < dm. When the focal radius aY is decreased, then
k=N,/N,-l-exp(A/A,) (-A/A,at A<A,). the length of the region with high photon density becomes
shorter and the probability of conversion almost does not
(11) change.
Let us estimate A,,. At the conversion region the r.m.s. Let us consider another example: 2& - 1,~ I,. In this
radius of the laser beam depends on the distance z to the case A, is only slightly larger than A,, mti, but the laser
focus (along the beam) in the following way [3]: target has a relatively low density (and large length ( /3,>>
that is important to avoid multiphoton processes to be
rY=a,/m, (12) discussed in the next section. In this case the radius of
laser beam ry - aY along the target and the density of laser
where 6, = 2ma:/A, aY is the r.m.s. focal spot radius,
photons ny -A/(Tooatl,). The probability of conver-
and A is the laser wavelength. The expression for /3, is
sion p*ny(l;Zy=l at
valid for a Gaussian shape of the beam in the diffraction
limit of focusing. The density of laser photons lr rid
A,= 2 zTA 0,min= 251,[cm] J. (19)
A 2%
f2Y =-exp(-r’/r,Z)F~(~+ct),
2
=r,uO Here, as before, we assumed x = 4.8. With such a focus-
ing the angular divergence of the laser light is
/F,(z)dz=l. (13) (20)
Let us assume that the linear photon density is uniform The value of A, only slightly varies until the collision
along the beam: Fy = l/1, and /3, <( I,. The probability angle cro < (Ye. For a0 z=-(TV,A, is larger by a factor x

I. INTRODUCPORYPAPERS
10 V. Telnov/ Nucl. Ins@.and Meth. in Phys. Res. A 355 (1995)3-18

= GcY,,/cY, [3], although, in principle, using complicated 4.3. Lasers

optics one can get the same A, for any collision angle.
From two previous sections follows that to obtain the
conversion probability k = 65% (A = A,) at x = 4.8 a
4.2. Influence of a strong field on processes in the conver-
laser with the following parameters is requried:
sion region
Flash energy: A, - max(lQ, [cm], 4Ee [TeV], J;
Duration: ct = max(Z,, 0.17 E [TeV]), cm;
In the conversion region the density of laser photons Repetition rate: n bunches X reprate of a collider;
can be very high, which leads to multiphoton processes Wavelength: A = 4.2 E, [TeV] p,rn or w, = 0.3/
[18,19]: E,, [TeV] eV;
e+ny+e+ny,, y+n-ra+e+e-, Angular divergence: near the diffraction limit.
The first numbers for flash energy and duration are deter-
where e, y are a high energy electron and photon respec- mined by diffraction, the second by nonlinear effects in the
tively, and yu denotes a laser photon. Nonlinear effects are conversion region. For example, at E, = 0.25 TeV and
described (partially) by the parameter 1, = 200 pm (SLAC, KEK) a laser with a flash energy
A - 1 J, I, N 400 p,rn and A = 1 pm is required. The first
eFh two numbers are determined by nonlinear effects. For
(= - (21)
rn0+ ’ TESLA with 1, = 2 mm a laser with A - 3 .I and I, u 0.2
mm is required. Here, nonlinear effects are not essential.
where F is the field strength (E,B) and w,, is the photon Obtaining Joule pulses of picosecond duration is not a
energy. At 6 -Z 1 an electron interacts with one photon problem for modem laser technique. The main problem is
from the field (Compton scattering), while at .$B 1 an a high repetition rate. Nice reviews on achievements and
electron feels a collective field (synchrotron radiation). perspectives of conventional lasers for photon colliders
What values of 5 are acceptable? In a strong field were given by Geissler [20] and Meyerhofer [21]. Solid
electrons have a transverse motion, which increases the state lasers with a few joule flash energy and picosecond
effective mass of the electrons [18]: m2 + m2(1 + 5’). duration have a repetition rate of only about a few Hz.
Due to this fact at 6 2 = 0.3 (and n = 4.8) y, decreases by This is connected with overheating of the media. For
5%. example, a Nd:glass laser (A = 1.06 Frn used now in
What is the value of 5 in the our case? Assuming experiment E-144 at SIAC [21] has 2 TW power and 1 ps
2& 5: I, (the case of a low density target, see previous duration. It is exactly what is needed for the collider with
subsection) and taking into account that the field in the 2Eu = 500 GeV energy. However, its repetition rate is
laser focus is F2 = 4A/a$, and A = M, (Eq. (19)) one only 1 Hz. Replacing pumping flashlamps by diodes
gets [7,81 (lasers) will substantially reduce the thermal problem and
the repetition rate can be increased up to 100 Hz [21]. An
2 %A
5 _---rra UC 1, k.
2,

(22) array of lo-100 slabs will provide a repetition rate suffi-

cient for photon colliders. Gas lasers have also good
parameters and perspectives.
At x = 4.8(uc/g0 = 0.75 and A = 4.2Eo(TeV) p,rnl and The success in obtaining picosecond pulses is con-
k = 1 we get nected with a chirped pulse technique (“chid” means
time-frequency correlation in the pulse). After amplifica-
6’ = O.OSE,[TeV]/I,. (23)
tion a long chirped pulse is compressed using a grating
For example, if I, = I, = 200 pm and E, = 1 TeV, then
t2 = 2.5. This is not acceptable. To decrease 5, keeping
k = const, one has to use a longer photon bunch, i.e. I, GRATING PAIR
must be longer than I,. From Eq. (23) it follows that, if
6 ’ = 0.3, then

I, = O.l7E,[TeV] cm. (24)

sHoRr WHITE
Substituting I, instead of 1, into Eq. (19) we get PULSE

A, = 4E,[TeV] J. (25)
For large E. and short bunches this requirement to the
LONGCHlRPEOPULSE
energy A, is stronger, than follows from the simple con-
sideration of the conversion probability, Eqs. (161 and
(18). Fig. 11. Compressionof a chirped laser pulse using grating pair.
 V. Telnov/Nucl. Instr. and Meth. in Phys.Rex A 355 (1995) 3-18 11

pair to picosecond duration (see Fig. 11). Stretching and - conversion of photons into efe- pairs in ye and yy
compression by a factor of 1000 has been demonstrated. collisions (coherent pair creation);
Very attractive lasers for photon linear colliders are - beam displacement in ye collisions.
free-electron lasers [5]. They have a tunable wavelength These effects determine attainable yy, ye luminosities. Let
and a high repetition rate. However, it will be difficult to us consider them.
generate joules in 1 ps. The task becomes much simpler, if
the FEL generates a long chirped pulse, which is com- 5.2.1. Coherent pair creation in the beam field
pressed after that by a grating pair [8]. The probability of pair creation per unit length by a
Analyses carried out by FEL experts [22-251 show that photon with energy w in a magnetic field B ( 1B 1+ 1E 1)
there are no (very) serious problems in the way of FEL for for our case) is [7,8,28,29]
photon colliders (except money) and they should be devel- a2 B
oped right now together with linear colliders. CL(K) = F gT(‘+>
e 0
where
5. Collision effects, ultimate luminosities of y y, ye
w B
collisions KG --
mc2 B,’
5.1. Collision schemes
m2c3
B,= - = 4 4 x 1013 G
It is well known that for e +e- linear colliders collision eh .
effects (beamstrahlung and beam instability) impose strict is the the critical field, and re = e2/mc2 is the classical
restrictions on beam parameters and determine the attain- radius of the electron.
able luminosity. At first sight in ye and yy collisions at
least one of the beams is neutral so that collision effects T(K) = 0.16~-%,2/,(4/3~),

are absent. This is not the case. Let us first consider how = 0.23 exp( -8/3K) for K -+z 1,
beams are colliding. Three schemes are discussed:
Scheme A: The conversion region is situated close to = 0.38K-“3 for K%’ 1. (27)
the interaction point (i.p) at a distance b N 2uz, and after In our case w _ E,; therefore one can put
conversion all particles travel directly to the i.p. [8,26].
Km r= yB/B,. (28)
Scheme B: After conversion at some distance b from
the interaction region the particles pass the region with the The probability to create e+e- pair during the collision
transverse magnetic field, where “used” electrons are time is
swept aside in the horizontal (x) (or vertical (y)) direction
by a distance larger than a; [3,7,8]. Thereby one gets
more or less clean ye or yy collisions.
Scheme C: After conversion electrons are defocussed
If we allow some pair creation probability p, then from
by a plasma lens. This idea was proposed at this workshop
this equation one can find the corresponding 2’. In Table 2
[27]. Here I will not consider this interesting approach.
the values of YP are presented for p = 0.05 and various uz
The first scheme is simple, but background conditions
and Ea. We provided 2’ with the index p to designate
are much worse (mixture of yy,ye,ee collisions, larger
effect of “pair creation”. One can see that for existing
disruption angles), although, for yy collisions at a proper
projects r, - 1.
choice of beam parameters, the ye,ee background is sub-
stantially suppressed owing to beams repulsion. Below we
52.2. Radiation energy losses of electrons fbeamstrah-
will consider both schemes A and B and compare attain-
lung) in ye collisions
able luminosities.
The field of the opposing beam leads to beamstrahlung.
The relative energy loss of an electron in field B per
5.2. Beam collision effects

During beam collisions electrons and photons are influ- Table 2

enced by the field of the opposing electron beam. In the Values of T, for p = 0.05
case of yy collisions the field is created by “used” u,,. cm E,, TeV
electrons deflected after conversion by the external field
0.25 0.5 1.0
(not deflected in the scheme A). In ye collisions the field
0.01 1.32 1.43 1.67
is also created by the main electron bunch used for ye
0.05 0.74 0.9 1.2
collisions. A strong field leads to:
- energy spread of the electrons in ye collisions; 0.1 0.62 0.74 0.91

I. INTRODUtXORY PAPERS
12 V. Telnov / Nucl. Instr. and Meth. in Phys. Res. A 355 (1995) 3-18

Table 3 where x0 is the deflection of electrons with the energy E,.

Values of T, for 6 = 0.05 The field inside the “main” electron beam
a,, cm I?,, TeV eN
0.25 0.5 1.0 B,- - (36)
a;%
0.01 0.19 0.33 0.52
0.05 0.07 0.11 1.16 From Eq. (36) and connection YP= yB/B, we get
0.1 0.048 0.07 0.11 eN W’ N?y
u=-=-=-,
X (37)
“I& q BoTP ffazTp

collision length - 0; is [7,8,30,31] where YP is given in Table 2.

To provide a deflection x0 in the external field B, it is
necessary to have a distance between the conversion re-
(30) gion and the i.p.

where T was defined above and 4N-y*er,( 1 - OX)

b2=-= 2-Qlxo (38)
F(r) = 1/(5T+ 1) for Y< 100. eBe aB, q’l-‘,, ’
(31)
Value of 6 is restricted usually to the level l-10%. For a where ‘Z’, is given in Table 3.
given 6 one can find T, (index b is “ beamstrahlung”). In
Table 3 the values of ?‘,, are presented for 6 = 0.05 and 5.3. Ultimate luminosity in ye collisions
various uz and Eo.
It is of interest that, if S is small, then the pair creation 5.3.1. Scheme A (without deflection)
probability is also small. There are three collision effects here: beamstrahlung;
pair creation; and beam-beam instabilities.
5.2.3. Beam-beam displacements As was pointed out in the previous section, if beam-
In Scheme A of ye collisions (without deflection) the strahlung losses are small, then the pair creation probabil-
main electron bunch collides with the opposing “used” ity is also small: p < 6 -=K1. The other two points are the
electron bunch. Let us assume for simplicity, that all same as for ee collisions. Therefore
“used” electrons have energies equal to E,. The problem L ye,max- KY,. (39)
of instabilities of e+e - and e-e- beams is well known
[32]. If e-e- luminosity drops, then L,e will go down as
well. The maximum attainable ee luminosity due to this 5.3.2. Scheme B (with deflection)
effect The maximum ye luminosity is

W2f
(32) L
-F
-
4T( b/y)al; .
(40)
where f is the coIlision rate. L,, is only smaller. Here the horizontal beam size is determined by pair cre-
If “used” beam is deflected (Scheme B) after conver- ation while the distance b is determined by two effects:
sion by the distance xr,, then due to repulsion the main beamstrahlung and beam displacement.
bunch is displaced during collision by a distance Taking gX from Eq. (37) and b from Eq. (38)
(beamstrahlung) we get
eBuzz cTzr,N
Ax-----. (33)
Eo YXO
Lqe - (41)
The luminosity is not lost if Ax < uX, so when
mzrCN
x0> -. (34)
Y?z
Table 4
5.2.4. Beam fiehis, defection The coefficient C in F!q. (42)
The field produced at the i.p. by “deflected” (by an u,, hds E,,, TeV
external field) electrons of the converted beam is approxi- 0.25 0.5 1.0
mately equal to (81
0.01 0.46 0.31 0.23
2eN(l - OSk) 0.05 1.7 1.3 1.04
B,=IEI+IBI- 9 (35) 0.1 3.4 2.4 1.84
%X0
 V. Telnov/Nucl. Instr. and Meth. in Phys.Res. A 355 (1995) 3-18 13

For k = 0.65 (A =A,,), B, = 30 kG and values of rP, rt, Table 6

from Tables 2 and 3 we get Ultimate L,,, 1O34cm-*s-l at 2E, = 500 GeV in Scheme A
(without deflection)
4” =
dg) f[Hz] x 1030x c cm2ss1, (42)
NLC
1.5
DLC
7.5
VLEPP
3.5
TESLA
30
where values of C are given in Table 4. With +20%
accuracy Eq. (42) is described by
(see one paragraph above). The minimum vertical beam
a,[cm]f[Hz] X 1031
size at the i.p. is
\IE,[Tevl cm-*s-r. (43)
mYy”b/-r- 21,/y> (48)
This is the ultimate L,, due to pair creation and beam- where 1, is given in Section 4.3. For electron beams with
strahlung. If x0 (and b) are determined by the beam a, = 0.1-0.75 mm and E = 0.25-l TeV, bmin varies be-
dispalacement during collision and a,, as before, by pair tween 0.1 and 0.35 cm and uY = 2-6 nm. These numbers
creation, then we get are reasonable (see Table 1).
The yy luminosity in this scheme [8]

N2k2f Nk2auJp f
L YYNpN
4Po;UY 4 P rayuY
- 1.3 x 1030 E’/*[TeV]f[Hz] cm-*s-l.
k%+mlf [*I 3.7 x 1o32 cm-2s-l

(44) uY[nm] Ea [TeV]

Here we put 7-r = 1; that is good approximation. (49)
A joint limit on ye luminosity is
Taking !$ from Table 2 and assuming uY - 5 nm, k = 0.65
L,, = min(L,, L2). (45) we get the estimate for the yy luminosity in scheme B
Beam displacement is essential (L, < L,) at (without beam deflection) (see Table 6).

5.4.2. Scheme B (with deflection)

- ~ <15. (46) The deflection x0 and the distance b are determined by
pair creation (see Eqs. (35) and (38)). The minimum
For different projects Lye,max is given in Table 5. We
vertical size of the photon beam at the i.p is a., = b/y;
see that L,, is good for 2Eo = 0.5 TeV, but there will be
therefore the maximum yy luminosity [7, 81
problems for higher energes.
Beams in ye collisions should be flat. From Eq. (37) at N2k2f Nk2uzB& f
L - (50)
Yr N 0.5 one gets
” 4v(b/y)* = 161v,3B,(l - 0.5k) ’

where Tp - 1 (see Table 2). For k = 0.65 (A =A,) and

(47)
B=30 kG
For N = lo*‘, a, = 0.01 cm we get a, = lo-’ cm at
L yy - 2 + f [Hz]u,[cm]T,
X 103’ cm-*s-l. (51)
2E, = 0.5 TeV that is close to ox in e+e- collisions.
( 1

5.4. Ultimate luminosity in yy collisions The results for E, = 0.25 TeV are presented in Table 7.
In this table the yy luminosity per one collision, the
X4.1. Scheme A (without dejlection) photon spot size at the i.p., the distance between the i.p.
In this scheme the electron beams must be flat. The and the conversion region and the deflection of the con-
horizontal size 0; at the i.p. is determined by the coherent
pair creation (Eq. (37)); it is the same as for ye collisions Table 7
Ultimate L,,, 10% cm-2s-1 at 2E, = 500 GeV in Scheme B
(with deflection)
Table 5
Ly’y,10% L,, /CO& 103’ ayr urn b, cm x0, nm
Ultimate LYc, 1O34cm-2s-1
NLC 3 1.9 10 0.5 40
2E,,TeV NLC DLC VLBPP TBSLA
DLC 12 14 9 0.5 40
0.5 0.6 0.9 0.3 2.0 VLEPP 6 200 25 1.3 300
2 0.3 1 0.2 3.0 TESLA 50 60 12 0.6 60

1. INTRODUCTORY PAPERS
 14 V. Telnov/ Nucl. Instr. and Meth. in Phys. Res. A 355 (I 995) 3-18

verted beam (for E = E,) are also given. In this estimation beam. In the head-on collision of Bat beams soft electrons
we assumed that the electron beam size at the i.p. is acquire the angles [7]
smaller than a,, -b/y, which is about lo-20 nm (see
Table 7). This size is comparable or even larger than ay
for designed e+e- colliders (Table 11, but much smaller
than a,. Damping rings used usualIy as injectors produce
For E,,,, = 0.024,
naturally flat beams and there is a problem in obtaining
beam sizes small in both directions. l/.7
N/10”
Comparing Tables 6 and 7 we see that in both schemes ll.$- 3 mrad.
a,[mmlG[TeVl
with and without deflection the ultimate luminosities are
comparable. Each scheme has its advantages and disadvan- For NLC, TESLA, DLC at 2E,, = 500 GeV, ab N 15 mrad,
tages. while for the 100 GeV photon collider considered in Ref.
There is a third intermediate scheme of collisions. As [26]: E, = 100 GeV, N = 4 X lo”, a, = 0.75 mm, the
soon as electron beams are naturally flat, then to avoid disruption angle 8, N 70 mrad. This estimation describes
coherent pair creation one can set Tp,by varing a,, and Scheme A (without deflection) and the “intermediate”
make a small separation in the vertical direction to reduce scheme (with a small vertical deflection).
the background of ee and ye collisions. This case is close In Scheme B with a “large” horizontal deflection (to
to Scheme A, but beam repulsion occurs by an external obtain TP _ 1) particles with smaller energy come to the
field. At b = 0.5 cm the magnetic field of 10 kG separates i.p. with a larger displacement and therefore get a kick in
250 GeV electron bunches by a distance of 30 nm. Resid- the field of opposing beam which is much smaller than in
ual ee collisions are no problem; due to beam repulsion ee a head-on collisions.
luminosity will be almost negligible at any reasonable The results of MC simulation of beam collisions in
beam parameters. This separation is necessary only for different schemes are presented in the next section.
suppression of the residual ye luminosity.

6. MC simulation of y y , ye collisions
5.5. Disruption angles
As we have seen, the picture of collision is so compli-
After conversion the electrons have a wide energy cated that the best way to see a final result is the simula-
spectrum. Estimations [7] and simulation (next section) tion. This program, written by the author only a month
show that at k = 0.65 sume electrons have an energy as ago, takes into account the following processes:
low as 2% of the initial energy. During a beam collision 1) Multiple Compton scattering in the conversion re-
these electrons get kick angles in the field of the opposing gion. It was assumed (as a first approximation> that polar-

Table 8
Parametersof beams and results of simulation; the notation a,b . f correspond to (a), (b) . . . (f) in Figs. 12 and 13; the two right columns
correspond to Fig. 13
a b C d e f a b

Type of collisions YV vv YV vv VY vv Ye Ye
# conversion lengths 1, 1 1, 1 1, 1 1, 1 171 1, 1 LO LO
2E,, GeV 500 500 500 500 500 200 500 500
N/1Ct’o 5 5 5 5 5 40 5 1
q, nm 20 6.1 150 150 150 200 50 100
ay, nm 20 6.1 20 5 5 4 12 12
a,, mm 1 1 1 1 1 0.75 1 0.1
b, cm 1 1 2 0.35 0.35 0.1 3 2
B,, kG 0 0 0 0 20 0 0 0
B,, kG 30 30 30 0 0 0 30 30
L(geom), 103’ 48 530 6.5 27 27 1600 22 0.65
L,,, 1030 10 18 1.7 13 9 750 2.7 0.07
L,, (z > 0.6) 5.2 12.8 0.63 2.8 3 165 < 0.003 -0
L,,, 103o 0.1 0.07 0.017 4.8 0.7 60 3.4 0.28
L,, (I > 0.6) -0 -0 -0 1.6 0.33 13 2.7 0.21
4 x, m8x,mrad 4.5 4.2 4.2 6 3 55 4.5 2.5
9 y, max.mrad < 0.5 < 0.5 < 0.5 7.5 6.6 42 6 3.6
 V. Telmv / Nucl. Instr. and Meth. in Phys. Res. A 355 (199.5)3-18 15

a b

%t; -

c:)
1
0
0 0.25 0.5 0.75
!l

2.5

2-
0

r
ij_ll
i’

0.25 0.5 0.75

7.
e)
-
1
,

0’
0
’
0.25
. ’
0.5
‘.
0.75
L .!.

f)

1.5 .

1 -

0.5 .
,:;,.;;,:.....__. :,c.., :,
o _r_ , ~:‘)_.~-.:.-Lcr.... .._.~ _ci ___
0
--‘.‘*---.,r,: __, ,.__, __: _.i’>
0 0.25 0.5 0.75 1 0 0.25 0.5 0.75 1
z-W&E, z-W/2&
Fig. 12. Spectral luminosities for yy collisions obtained by simulation taking into account main effects at conversion and interaction
regions. Dashed histogram is ye luminosity normalized to yy luminosity. Parameters of the beams and other results are presented in Table
8.

a b7

0 0.25 0.5 0.75

z=w,12Eo

Fig. 13. Spectral luminosities for ye collisions obtained by simulation taking into account main effects at conversion and interaction regions.
Dashed histogram is ‘yy luminosity normalized to ye luminosity. Parameters of the beams and other results are presented in Table 8.

I. INTRODUCTORY PAPERS
16 V Telmv / Nucl. Instr. and Meth. in Phys. Res. A 355 (1995) 3-18

ization of the electrons does not change in Compton o-----------

scattering.
/ r___,___________

2) Deflection by an external magnetic field and syn- 3

chrotron radiation in the region between the conversion

region and the i.p.
3) Electromagnatic forces, coherent pair creation and
beamstrahlung at the i.p.
Beams are composed of about 1000 macroparticles
(strings of 0.3~~ length). As output the program simulates lm
luminosity spectra, energies and angles of final particles. f l

The results are presented in Table 8 and in Figs. 12 and

13. The dashed histograms in the figures represent the ye
and yy background to yy, ye collisions respectively.
The considered cases have the following distinct fea-
tures:
Fig. 12a: yy round electron beams with a, = uY= b/y, Fig. 14. A schematic layout of the interaction region in a photon
collider.
deflection by external field;
Fig. 12b: as Fig. 12a but a, = uY= 0.3b/ y;
Fig. 12~: yy flat beams, uY = 0.5b/y, horizontal deflec-
tion; ity is the same as in head-on collisions. The characterictic
Fig. 12d: yy flat beams, small b, no deflection; disruption angles for a photon colliders are about 10 mrad
Fig. 12e: as Fig. 12d but with vertical deflection; (see Table 8), but can reach 60 mrad for a 200 GeV
Fig. 12f: yy [26]: 2E, = 200 GeV, flat beams, very large collider with superluminosity (case f)).
N, no deflection; A very powerful laser is focused near to the i.p. To
Fig. 13a: ye with deflection, N = 5 X 10”; obtain maximum conversion probability the focusing mir-
Fig. 13b: ye with deflection, N = 10”. rors should be placed at zero angles. In this case they
Some comments on the results shown in Figs. 12 and should have holes for passing the initial electron beams
13: and for removal of the highly disrupted beams from the
1) The wide enhancement around z = 0.25 in Figs. i.p. The angular divergence of the laser light (see Eq. (201)
12a-12c is due to collisions with soft photons that oc- is
curred in the multiple Compton scatterings.
A
2) In Fig. 12d there is a very large ye background; it is
substantially suppressed in Fig. 12e where vertical mag-
netic deflection is used.
% ’
J -
nJ,
-0.02.

Mirrors can reflect a flux of about 1 J/cm* in ps region

3) The collider corresponding to Fig. 12f produces
without damage. From the equation
fantastic yy luminosity, but disruption angles are large
and the number of reactions yy -+ hadrons per beam 35
A/cm2 - - 1 J/cm* (55)
crossing is 300!
“(%)2
4) The ye luminosity in Fig. 13 at low t is due to
collisions of Compton photons with beamstrahlung pho- we find a focal distance F - 50 cm.
tons, The low energy yy luminosity is produced by soft In the scheme with magnetic deflection of the con-
photons created in multiple Compton scatterings and verted electrons a small thin magnet with the field of
beamstrahlung photons. lo-30 kG should be placed near the i.p. This problem was
Other observations and conclusions can be made by the discussed at this Workshop [17] and needs a further study.
reader, analyzing Table 8 and Figs. 12, 13.

8. Measurement of y y luminosity
7. Interaction region
A system produced in a yy collision is characterized
A sketch of a possible set-up around the i.p. is shown by its invariant mass WYy = 46 and rapidity q =
in Fig. 14. As in efe- colliders, in a photon collider 0.5 ln(w,/oa). We should have a method to measure 1)
beams are supposed to collide using a “crab crossing” d*L/dWdq and 2) A,, Ay2 or, in other words, d L,/dWdq
scheme [33], so that after the collision the disrupted beams and dL,/dWdv (0, 2-total helicity of the system). These
travel outside lenses. The electron beams in this scheme can be measured using the process yy + efe-[34]. The
are tilted with respect to the beam direction at some angle cross section of this process can be found elsewhere
ac equal to half of the crossing angle so that the luminos- 116,351.
 V. Telnov /Nucl. Instr. and Meth. in Phys. Res. A 35.5 (1995) 3-18 17

When p + 1, then o,,o/oz - m2/s (excluding the re- respects is complimentary to that in e+e- collisions. In
gion of small angles). Therefore, the measurement of this these proceedings one can find the detailed analysis of
process will give us dL,/dzdq. How to measure physics at photon colliders. Below one can find a list of
dL,/dzdq? This can be done by inversion of the helicity some reactions, which are obviously very interesting:
(A,) of the one photon beam by means of changing ye + WV;
simultaneously the signs of the helicities of the laser beam
used for e --) y conversion and that of the electron beam. ye + Z’e;
In this case the spectrum of scattered photons does not ye+e* +ye;
change while the product A,, A,, changes its sign. In other
words, what was before L, is now L,, which we can ye + ?S 4 qqe;
measure. The cross section for this process cr(]cos 8 I< yy -+ w+w-;
0.9) - 10-36/s[TeV2] cm2. This process is very easy to yy + H+H-;
select due to a zero coplanarity angle.
Instead of e+e- pairs, muon pairs also can be used yy + L+L-;
[36]. They have the same properties and cross sections. I y-y + Ho.
suggested in Ref. [34] to use efe- pairs only because at
high energies the energy of the electrons can be measured
much better (using a calorimeter) and easier than for
muons.

9. Luminosity and hadronic background

In designs of linear colliders a luminosity of

2

L= 1034 +& cm-2s-1

( 1 0 2.5 5 7.5 10 12.5 15 17.5 20

is usually assumed as a reference point. In the current ,-

designs the effective collision rate is 103-104/s. It is
achieved by using bunch trains of order nb - 100 bunches 0.6 -

(see Table 1). In NLC, for example, the bunch spacing is

only 1.4 ns and it is not simple to distinguished each
0.4 -
crossing in time. Let us assume that it is possible and take
as a reference a luminosity/crossing about 103’ cm-2s-1.
One of the potential problems for photons colliders is 0.2 -
hadronic background. At the investigated energies a(yy
.,__e’e-+y’+F’F
+ hadrons) = 300 nb and it is expected to be about 500 nb
---...._._._. . . -.
at TeV energies. At reference L/co11 = 103’ we get 0.5 O- I..I..I..I..l-..T..;

0 2.5 5 7.5 10 12.5 15 17.5 20

event/crossing. In general, particles in this process have
small p I , but hard QCD processes also contribute to the
total cross section. The analysis of this problem [37,38]
shows that this background is not very serious at such a
luminosity per collision. However, it seems rather prob-
lematic to work at luminosity 1033/collisions [26]. The
7.5~
5 -

analysis of collision effects given in previous sections

shows that L = 103’/crossing is not a limit for photon
colliders. What the limit imposed by hadronic background 2.5 -

is, is not yet clear and further analysis is required. -.

' -. ._ezg+y*+ W’W-
-... ---....._.____
0 ,,‘..‘,.‘..‘.,‘.,‘..‘...
0 2.5 5 7.5 10 12.5 15 17.5 20

10. Physics (some examples) 1/X&/4M1

Fig. 15. Comparison of cross sections for charged pair production
The number of papers on physics at ye, yy colliders is in e+e- and YY collisions. The cross section (T =
growing exponentially in the last three years, proving that (~a’/M*)f,(xk p = S (scalars), F (fermions), W (W-bosons);
the physics is very rich and interesting here and in many M is particle mass; x = WG/4M2, fp(x) are shown.

I. INTRODU~ORY PAPERS
18 V. Telmv / Nucl. Instr. and Meth. in Phys. Res. A 355 (1995) 3-18

Photon colliders are a factory of Z” and W * bosom [13] I.F. Ginzburg, Proc. 9th Int. Workshop on Photon-Photon
and their properties and interactions can be studied here Collisions, San Diego 1992 (World Scientific).
very well. [14] S. Brodsky, Proc. Workshop on Physics and Experiments at
In ye collisions one can produce a supersymmetric Linear Colliders, Hawaii, 1993 (World Scientific).
[15] G. Low and B. Wiik, Proc. 5th Workshop on Next Linear
particle almost twice heavier than in e+e- collisions,
Colliders, SLAC, Stanford, Oct. 1993, SLAC-436.
because the partner is a light particle.
[16] T. Barklow, SIAC-PUB-5364 (1990).
In yy collisions a resonance H&s production is of [17] F. Richard, these Proceedings (Workshop on Gamma-
particular interest [16,10]. SUSY predicts one of the Gamma Colliders, Berkeley, CA, USA, 1994) Nucl. Instr.
Higgs’s with a mass below 130 GeV. Simulation shows it and Meth. A 355 (1995) 92.
can be studied at photon colliders. [18] L. Landau and E. Lifshitzs, Kvantovaya mekhanika, vol. 1,
It is remarkable that cross sections for the production of (M. Nauka).
charged pairs in yy collisions are higher than in e+e- [19] I. G&burg, G. Kotkin and S. Polityko, Yad. Fiiika 40
collisions (see Fig. 15). (1984) 1495; 37 (1983) 368.
[20] K. Geissler, presented at the Workshop on Physics and
Experiments with Linear Colliders, 1991, Lapland, Finland,
unpublished;
Acknowledgements See also Ref. [B];
Talk at ECFA Workshop on e+ e- Linear Colliders, LC92,
I would like to thank A. Sessler for organizing the First Garmish-Partenkirchen, 1992, unpublished.
Workshop on Gamma-Gamma Colliders, which I hope [21] D. Meyerhofer, these Proceedings (Workshop on Gamma-
Gamma Colliders, Berkeley, CA, USA, 1994) Nucl. Instr.
will stimulate not only theoretical but experimental works
and Meth. A 355 (1995) 113.
as well. I am very grateful to I. Ginzburg, G. Kotkin, V. [22] E. Saldin et al., Proc, and Workshop on Physics at VLEPP,
Serbo, and P. Chen for joint work and to V. Balakin, D. Protvino, 2-4 June 1992, vol. 2, p. 96.
Burke, D. Cline, K. Geissler, W. Kozanecki, D. Meyer- [23] A. Sessler, presented at the Workshop on e+e- Linear
hofer, D. Miller, V. Parkhomchuk, B. Palmer, F. Richard, Collisers, Garmish-Partenkirchen, July 1992, unpublished.
J. Spenser, A. Skrinsky, and W. Vernon for useful discus- [24] E. Saldin, V. Sarantsev, E. Schneidmiller and M. Yurkov,
sions. these Proceedings (Workshop on Gamma-Gamma Colliders,
Berkeley, CA, USA, 1994) Nucl. Instr. and Meth. A 355
(1995) 171.
[2.5] S. Hiramatsu et al., ibid., p. 133.
References [26] V. Balakin and A.A. Sery, ibid., p. 157.
[27] S. Rajagopalan, D. Cline and P. Chen, ibid., p. 169.
[I] I. Ginzburg, G. Kotldn, V. Serbo and V. Telnov, Pizma [28] T. Erber, Rev. Mod. Phys. 38 (1966) 626.
ZhETF 34 (1981) 514; JETP Lett. 34 (1982) 491 (Preprint [29] P. Chen and V. Telnov, Phys. Rev. Lett. 63 (1989) 1796.
INF 81-50, Novosibirsk (1981) in English). [30] R. Noble, Nucl. Instr. and Meth. A 256 (1987) 427 and
[2] C. Akerlof, Preprint UMHE 81-59, Univ. of Michigan (1981). references therein.
[3] I. G&burg, G. Kotkin, V. Serbo and V. Telnov, Nucl. Instr. [31] U. Amaldi, CERN-EP/87-169, Lecture at US-CERN School
and Meth. 205 (1983) 47 (Preprint INP 81-92, Novosibirsk on Particle Accelerators, South Padre, Texas, 1986.
(1981) in English). t321V. Balakin and N. Solyak, Preprint INF 82-123, Novosibirsk
[4] V. Balakin and A. Skrinsky, Preprint INF 81-129, Novosi- (1982); Proc. 8th All Union Workshop on Char. Particle
birsk (1981); A. Skriusky, Uspekhi Fiz. Nauk 138 (1982) 3. Accelerators Dubna, 1983, vol. II, p. 263; Proc. 8th Int.
[5] A. Kondratenko, E. Pakhtusova and E. Saldin, Dokl. Akad. Conf. on High Energy Accelerators, Novosibirsk, 1987, p.
Nauk 264 (1982) 849. 151.
[6] I. Ginzburg, G. Kotkin, S. Panfil, V. Serbo and V. Telnov, t331 R. Palmer, SLAC-PUB 4707 (1988).
Nucl. Instr. and Meth. 219 (1984) 5. [341 V.I. Telnov, Proc. Workshop on Physics and Experiments at
[7] V. Telnov, Nucl. Instr. and Meth. A 294 (1990) 72. Linear Colliders, Hawaii, 1993 (World Scientific).
(81 V. Telnov, Proc. Workshop on Physics and Experiments with L351K. Ispirian et al., Yad. Fiz. 11 (1970) 1278; Sov. J. Nucl.
Linear Colliders, Lapland, Finland, September 9-14, 1991. Phys. 11 (1970) 712.
[9] F.R. Arutyunian and V.A. Tumanian, Phys. Lett. 4 (1963) [361D. Miller, these Proceedings (Workshop on Gamma-Gamma
176; Colliders, Berkeley CA, USA, 1994) Nucl. Instr. and Meth.
R.H. Milbum, Phys. Rev. L&t. 10 (1963) 75. A 355 (1995) 101.
[lo] D. Borden, D. Bauer and D. Caldwell, SLAC-PUB-5715, [371T. Barklow, P. Chen and M. Peskin, Proc. Workshop on
UCSD-HEP-92-01. Physics and Experiments at Linear Colliders, Hawaii, 1993
[ll] Proc. 9th Int. Workshop on Photon-Photon Collisions, San (World Scientific).
Diego, 1992 (World Scientific). [381P. Chen et al., these Proceedings (Workshop on Gamma-
[12] Proc. Workshop on Physics and Experiments at Linear Col- Gamma Colliders, Berkeley, CA, USA, 1994) Nucl. Instr.
liders, Hawaii, 1993. and Meth. A 355 (1995) 107.