This can be very long - for example, 9.9 hours in a full cir-
cumference eRHIC electron ring! Further, it is not possible
to accelerate (or decelerate) electrons through intrinsic spin
resonances, which are located at energies given by
Eresonance = J 0.441 [GeV] (16)
where .J is an integer. In this case full energy electrons
must be injected pre-polarized. It is natural to consider us-
ing permanent magnet technology for some or most of the
ring lattice magnets in such a fixed energy ring. One way to
inject full energy pre-polarized electrons is to use a full en-
ergy linac equipped with a polarized source. Another is to
use a conventional booster ring. Equation 15 shows that an
eRHIC booster with I T dipoles (p = 33.4 m) and a pack-
ing fraction of 0.5 (C = 420 m) has a polarization time of
only Tp01 = 74 s. Such a booster would accelerate electron
bunches from (say) a 1 GeV injection energy to a 10 GeV
flat-top, and then hold them there for a couple of minutes,
before injection into the electron storage ring.

r -1
o -2


-100 -50 0 50
Distance 'east' from IP [ml



Figure 2: Plan view of an eRHIC IR with spin rotators.
Interaction Region optics - spin rotators. In some
cases it may not be necessary to provide longitudinally po-
larized electrons at the IP. In this case the interaction region
optics are optics are relatively attractive, because weaker
dipole fields are permitted, with lower synchrotron radi-
ation linear power loads. Polarized electron experiments
require the spin vector to be longitudinal at the IP, while
the spin vector is naturally vertical in the arc of a ring.
The transformation from vertical to longitudinal polariza-
tion (and back again) is achieved by including spin rotators
in the interaction region optics. Figure 2 shows a straw
man eRHIC interaction region layout, including spin rota-
tors [5]. Not shown in the plan view is a vertical drop of
ahnost 1 m, which puts the electron ring near the floor of
the tunnel in the arcs. The spin rotator dipoles may have
much higher fields than in the arcs - as high as B = 0.43

T in the straw-man eRHIC optics. These few dipoles have
much higher linear heat loads than the common arc dipoles.
The natural potential advantages of a linac-ring collider
stem from the single pass nature of the beam. Not only
do the electrons collide with the ion beam with the initial
properties of the linac (rather than the equilibrium values
of a ring), but it is also possible to exceed the multi-turn
dynamical limits of a ring. The linac generates a low emit-
tance beam with a low energy spread, leading to a small
collision point beam size with a relatively large beta func-
tion that simplifies the interaction point optics. The beam
is naturally round, as is the ion beam. If collisions are re-
quired at only one IP, then there is no need for a spin ro-
tator in the IR optics, since the electron beam polarization
vector can be prepared with the correct orientation close to
the source of the linac. A single pass collider can provide
a polarized electron beam energies over a relatively broad
range, while a storage ring must avoid spin resonances. It
should be possible to alternate the sign of the polarization
in a linac rapidly, at will. Linac-ring collisions increase the
maximum permissible value of 6, the electron beam-beam
parameter -the electron beam can be "destroyed" by beam-
beam forces, and still have its energy recuperated. Equa-
tion 4 shows that this allows the number of ions per bunch
N; and 3* to be increased, and permits smaller emittance
ion beams, attained for example through the use of electron
Energy recuperation. With typical average electron
beam powers of order 1 GW, the recovery efficiency of
an ERL must be very high in order to avoid excessive
power budgets. This requires the use of a superconduct-
ing linac. Energy recovery has already been successfully
demonstrated at the Jefferson National Accelerator Facil-
ity IR-FEL facility, albeit with low power, current, and en-
ergy (250 kW, 5 mA, and 50 MeV) [17, 18]. Several in-
dicators at the JLab IR-FEL place an upper limit on the
beam loss at 2pA, or ~ 4 x 10I, an extremely small
value [19]. In a high power electron-ion collider ERL. frac-
tional beam losses at this upper limit could be unaccept-
able, since they potentially give rise to hundreds of kW of
uncontrolled beam power losses. Very little power can be
lost at cryogenic temperatures. More work is required to
understand both the origin of ERL beam losses, and their
possible cures.
In order to avoid beam-beam collisions of the acceler-
ated and decelerated beam, the two beams must propagate
in the same direction of the linac [20]. Thus the trans-
verse optics at each end of the linac must deal with beams
of very different energies. The energy difference should
be no more than about a factor of 10 - or perhaps much
more [21]. Since the energy recuperation must go down
to a low energy, multiple stages may be required. A straw
man four stage scheme is shown conceptually in Figures 3.

4 4

Upcoming Pages

Here’s what’s next.

upcoming item: 5 5 of 5

Show all pages in this article.

This article can be searched. Note: Results may vary based on the legibility of text within the document.

Tools / Downloads

Get a copy of this page .

Citing and Sharing

Basic information for referencing this web page. We also provide extended guidance on usage rights, references, copying or embedding.

Reference the current page of this Article.

PEGGS,S.; BEN-ZVI,I.; KEWISCH,J. & MURPHY,J. ACCELERATOR PHYSICS ISSUES FOR FUTURE ELECTRON ION COLLIDERS., article, June 18, 2001; Upton, New York. ( accessed May 22, 2019), University of North Texas Libraries, Digital Library,; crediting UNT Libraries Government Documents Department.

International Image Interoperability Framework (This Page)