There are many applications of electron beams in accelerator facilities: for electron coolers, electron lenses, and electron beam ion sources (EBIS) to mention a few. Most of these applications require magnetic compression of the electron beam to reduce the beam radius with the goal of either matching the circulating ion beam (electron lenses and electron coolers) or increasing the ionization capability for the production of highly charged ions (EBIS). The magnetic compression of the electron beam comes at a cost of increasing share of the transverse component of energy and therefore increased angles of the electron trajectories to the longitudinal axis. Considering the effect of the magnetic mirror, it is highly desirable to produce a laminar electron beam in the electron gun. The analysis of electron guns with different configurations is given in this paper with emphasis on generating laminar electron beams.

Generating laminar electron beams has a long and rich history of ideas and approaches. Some applications require magnetic compression of the electron beam with the goal to increase its current density, like the Electron Beam Ion Sources (EBIS)1 and electron lenses for the proton colliders.2,3 Considering that in the process of electron beam compression a fraction of the longitudinal energy is transferred into the transverse plane resulting in increased electron trajectory angle and, that the magnetic mirror entrance cone of main solenoid has a limited value, it would be beneficial for the beam transmission to have as laminar initial electron beams as possible.

If some beam electrons have trajectory angles outside of the entrance cone of the magnetic mirror, these electrons will reflect from the mirror back to the cathode and will oscillate between the cathode and the magnetic mirror while slowly drifting outwards.4 Due to energy exchange within the beam, some of these oscillating electrons can acquire enough energy to hit the surface of the cathode or Wehnelt electrode and produce secondary electrons with multiplicity larger than 1 providing a positive feedback mechanism for a process of accumulating oscillating electrons.5 When the accumulated space charge of these electrons exceeds a certain threshold, the oscillated electrons can bunch, affecting the primary electron beam, which also becomes bunched with a frequency determined by the period of these axial oscillations. Such beam excitations have been observed in real systems.6,7

While electron beams for electron coolers are not subjected to high magnetic compression, and sometimes undergo magnetic expansion to achieve additional transverse cooling, highly laminar electron beams with low initial transverse electron temperature are still needed for effective cooling.

The largest contributor to the transverse electron energy is the electron gun optics with its combination of electric and magnetic fields in a cathode-anode gap.

An attempt to form an optimum electrostatic field in a magneto-immersed gun with 5 anodes was tried at CERN,8 with the main purpose of generating an electron beam for an electron cooler with a transverse energy lower than 1 eV. An analysis of the beam electrons generated with this gun reveals a highly non-uniform dependence of the transverse energy component on the magnetic field with several minima and maxima. The magnetic field corresponding to the most pronounced minimum at 0.70 kG was selected as a working point for this gun with an electron current of 8.3 A and energy of 60 keV. A sharp dependence of the transverse electron energy on the magnetic field was noted, which limited the effective magnetic field range.

As a trend, for most magneto-immersed electron guns, the radius of the Larmor electron motion decreases with increasing magnetic field in a gun region. This effect can be used to make the electron beam more laminar simply by increasing the magnetic field in the gun because the ratio of the Larmor radius to the beam radius is conserved at all magnetic fields for such guns. However, increasing the magnetic field on the cathode limits the achievable magnetic beam compression into the final solenoid, where the magnetic field is limited by technical issues. This limitation prompts the search for a solution which allows using lower magnetic field in the gun region and still produces a laminar electron beam.

For some time, an adiabatic electron gun with an almost uniform axial electric field in a cathode-anode gap9 has been considered being the best approach to generate a laminar electron beam. This electron gun has an anode, which overlaps the cathode, a feature providing the needed uniform electrostatic field distribution. The next development step for this gun was equipping it with a spherical convex cathode and with a Wehnelt electrode having a similar convex shape and acting as an extension of the cathode surface.10,11 This convex spherical shape of the cathode surface looks counterintuitive since the near-cathode electric field has visible angles with the guiding magnetic field. However, this magneto-immersed gun employing an electrode geometry with a convex cathode shape geometry provides smaller radius of the Larmor motion at all magnetic fields than a gun with a flat cathode. Simulations of guns with convex cathodes11 show a non-linear dependence of the transverse electron energy with the radial coordinate of its emission. A similar dependency of the transverse electron energy on the radius of emission from a cathode with an ellipsoid shape has been shown, with several maxima and minima within the full beam radius with a general trend of increasing the transverse energy with the radius of emission.

In Ref. 12, an analysis of some electron guns with non-adiabatic fields has been done and some dependencies of the beam laminarity on the gun geometry have been established. In this paper, we have performed further investigations of electron guns immersed in a magnetic field and the generation of electron beams using magnetic confinement. As before, for our simulations, we used a 2D software package from Field Precision LLC.13 

We analyzed different electron guns studying the dependence of the maximum value of the electron trajectory angle of all simulated particles emitted from the cathode as a function of the magnetic field at the cathode and the electron current. The simulations have been done for uniform magnetic and electrostatic fields inside the anode tube downstream of the cathode-anode gap, where the anode tube radius is constant and at an energy determined by the cathode-anode potential difference in a way that the Larmor radius remains constant in the area of analysis. For easy comparison of different guns, all angles are normalized to an electron energy of 25 keV. The result of a series of simulation for a specific electron gun is presented in a table where the maximum trajectory angles are listed as a function of the magnetic field and the beam current. The color-coded table of trajectory angles helps to visualize areas of magnetic field with small angles, where the gun performs best and areas with large angles to avoid. All electron gun simulations have been done for a cathode diameter of 9.2 mm, which is used in the BNL EBIS electron gun.

The schematic of an adiabatic gun10 built by Budker Institute of Physics for BNL EBIS is presented in Fig. 1. The electron gun has very smooth distributions of both axial and radial electrostatic fields. The geometry is quite popular because of this quality.

FIG. 1.

Budker Institute adiabatic electron gun with a plot of the axial distribution of the radial electrostatic field at radius r = 3 mm.

FIG. 1.

Budker Institute adiabatic electron gun with a plot of the axial distribution of the radial electrostatic field at radius r = 3 mm.

Close modal

Figure 2 presents the laminarity map for this adiabatic electron gun. Figure 3 presents a sub-selection of the data, i.e., the maximum angle for varying magnetic fields, but with a fixed electron current Iel of 12.0 A.

FIG. 2.

Dependence of the maximum electron trajectory angle in degrees on the magnetic field and electron current for the adiabatic electron gun shown in Fig. 1. The horizontal scale (top row) gives the magnetic field in kG, while the vertical scale (left column) lists the electron beam current in amperes.

FIG. 2.

Dependence of the maximum electron trajectory angle in degrees on the magnetic field and electron current for the adiabatic electron gun shown in Fig. 1. The horizontal scale (top row) gives the magnetic field in kG, while the vertical scale (left column) lists the electron beam current in amperes.

Close modal
FIG. 3.

Dependence of the maximum trajectory angle on the magnetic field for the Budker Institute adiabatic electron gun shown in Fig. 1 when operated at an electron current of 12 A.

FIG. 3.

Dependence of the maximum trajectory angle on the magnetic field for the Budker Institute adiabatic electron gun shown in Fig. 1 when operated at an electron current of 12 A.

Close modal

The scan of this map at electron current Iil = 12.0 A is presented in Fig. 3.

This dependence is smooth and does not have extremes. The common practice shows that such an electron gun is used with a magnetic field on the anode close to 2.0 kG and sometimes with magnetic fields as low as 1.5 kG.

The criteria for acceptable maximum angle are determined by the specific application. If the electron gun is to be used for generating a high-current, high-density electron beam using further magnetic compression with a superconducting solenoid having a maximum magnetic field of 6 T and the magnetic field on the gun as 1.5 kG, the acceptance cone is 9.06°. One can see that the operation of this gun seems safe at 1.4 kG, corresponding to a maximum angle of 7.2°, but with limited margin considering other contributors to the trajectory angle, like non-symmetrical gun electrodes, non-uniform electron emission, misalignment of the gun, and main solenoid.

At Brookhaven National Laboratory, we modified Budker’s gun. A cylindrical anode overlapping the cathode that we replaced with a conical funnel-like anode opened to the anode with angle of a cone 63°.12 We simulated a number of different configurations of the electron guns with electrostatic fields close to adiabatic. Most of these guns had a conical funnel-like anode and different cathode-anode distances, cone angles, and shapes. The dependence of maximum trajectory angle on the magnetic field for these electron guns is shown in Fig. 4.

FIG. 4.

Dependence of the maximum trajectory angle on the magnetic field for different geometries of electron guns with adiabatic electric fields. Electron current is Iel = 12.0 A.

FIG. 4.

Dependence of the maximum trajectory angle on the magnetic field for different geometries of electron guns with adiabatic electric fields. Electron current is Iel = 12.0 A.

Close modal

This plot demonstrates that these guns have individual dependencies of angle on the magnetic field similar to the one in Fig. 3 and none of them can provide 12 A electron beam with maximum angles smaller than 5° at magnetic field lower than 1.5 kG.

The Larmor motion of the electrons is determined by the external electric and magnetic fields in the cathode-anode gap of the gun in combination with the space charge field of the electron beam. An attempt was made to modify the external fields in a way that at some operating conditions the integral of all transverse actions on the beam electrons is smaller than that in an adiabatic electron gun. These field variations should be local, confined to the cathode-anode gap. Since the Larmor motion period is comparable to the length of these variations, we call such electron guns with sharply varying fields as non-adiabatic guns. The overall goal of this attempt is to generate a laminar electron beam (small trajectory angles) also for an electron gun with a magnetic field lower than that used at an adiabatic gun so that a higher magnetic compression of the electron beam is attainable.

A strong variation of the magnetic field in the cathode-anode gap can be produced with a magnetic shim made of soft iron located close to the cathode-anode gap and producing a local dip in the magnetic field distribution (Fig. 5).

FIG. 5.

Electron gun with non-adiabatic magnetic field.

FIG. 5.

Electron gun with non-adiabatic magnetic field.

Close modal

The electrostatic geometry of this electron gun is the same as used at the BNL EBIS and is a result of morphing from the adiabatic gun in Fig. 1. The laminarity map of this gun with the non-adiabatic magnetic field is shown in Fig. 6.

FIG. 6.

Dependence of the maximum electron trajectory angle on the magnetic field and electron beam current for the electron gun with the non-adiabatic magnetic field in a cathode-anode gap.

FIG. 6.

Dependence of the maximum electron trajectory angle on the magnetic field and electron beam current for the electron gun with the non-adiabatic magnetic field in a cathode-anode gap.

Close modal

One can clearly see regions with small trajectory angles, and one of them is located in a low-magnetic field area. For an electron beam current Iel = 12.0, the optimum magnetic field is 1.4 kG and the maximum trajectory angle at this field is 3.45°, which is significantly smaller than that for the adiabatic gun at this magnetic field. However, the price for the small angle at the optimum magnetic field is larger angles at lower electron currents, which means that during the current ramp the beam crosses areas with larger trajectory angles. This effect has to be taken into consideration when a pulsed electron gun is designed.

The magnetic field for a small-angle region on this map at a given electron beam current depends on the distance between the magnetically shimmed anode and the cathode (Fig. 7). This is a quite convenient dependence as the optimum magnetic field is reduced for shorter cathode-anode distances. In other words, the smaller the gap—and therefore the higher the gun perveance—the lower the optimum magnetic field.

FIG. 7.

Dependence of the optimum magnetic field on the cathode-anode distance for an electron gun with the non-adiabatic magnetic field operated at an electron beam current Iel of 12.0 A.

FIG. 7.

Dependence of the optimum magnetic field on the cathode-anode distance for an electron gun with the non-adiabatic magnetic field operated at an electron beam current Iel of 12.0 A.

Close modal

The effect of the magnetic field alternation in the cathode-anode gap depends on a thickness of the magnetic shim because it determines the value of absorbed magnetic flux. As one would expect, a thinner shim produces less pronounced effect than a thicker shim (Fig. 8). It is therefore possible to control the effect of the magnetic field variation on the beam trajectory angle by changing the thickness of the shim. Our simulation also demonstrated that if the magnetic shim is replaced with a magnetic coil in the same volume, the effect on the magnetic field distribution and on the gun performance can be made similar. The magnet coil has the convenience of controllable modifications of the gun performance in situ.

FIG. 8.

Dependencies of the maximum trajectory angle on the magnetic field for an electron gun as shown in Fig. 5 with two different magnetic shims, 5.1 mm and 2.5 mm thick. The electron beam current is Iel = 12.0 A in both cases.

FIG. 8.

Dependencies of the maximum trajectory angle on the magnetic field for an electron gun as shown in Fig. 5 with two different magnetic shims, 5.1 mm and 2.5 mm thick. The electron beam current is Iel = 12.0 A in both cases.

Close modal

The transverse electric field in a cathode-anode gap can be modified by enhancing the transverse electric field with the gun electrodes, which means either the Wehnelt electrode or the anode or both.12 The geometry of such a gun is shown in Fig. 9. The transverse electric field has a bump in the cathode-anode gap. Such modification of electrodes and their effect on the electron beam is controllable on the design stage. Our simulations show that the laminarity map of such a gun resembles the map of a magnetically non-adiabatic electron gun, but the areas of small trajectory angles extend to even lower magnetic field. As with magnetically non-adiabatic electron guns, the effect of the field modification depends on the strength of this modification. The stronger the modification, the more pronounced the contrast of the map, with smaller minimums and larger maximums in trajectory angles.

FIG. 9.

Model of an electron gun with a non-adiabatic electric field.

FIG. 9.

Model of an electron gun with a non-adiabatic electric field.

Close modal

The value of the optimum magnetic field for such a gun also depends on the cathode-anode gap (Fig. 10), but this dependence is opposite to the magnetically non-adiabatic gun.

FIG. 10.

Dependence of the optimum magnetic field on the cathode-anode distance for an electron gun with the non-adiabatic electrostatic field. The electron beam Iel is 12.0 A.

FIG. 10.

Dependence of the optimum magnetic field on the cathode-anode distance for an electron gun with the non-adiabatic electrostatic field. The electron beam Iel is 12.0 A.

Close modal

The optimum magnetic field for electron guns with the non-adiabatic electrostatic field for the geometry presented in Fig. 9 decreases with increasing cathode-anode distance. This effect may be explained by an increase in the period of the Larmor motion at lower magnetic field so that it matches the extended modified electrostatic field. For this kind of electron gun, the benefit of reducing the optimum magnetic field comes at the cost of a reduction of the gun perveance.

To estimate the contribution of the electron space charge to the pattern of the gun laminarity map, we simulated electron beam transmission with space charge and without it. The simulations have been done with identical magnetic and electrostatic fields for the non-adiabatic electron gun shown in Fig. 9. The electrostatic field in this simulation corresponds to the extraction of the electron beam with current 12.0 A. Figure 11 presents dependencies of the maximum relative Larmor radius as a ratio of the Larmor radius to the beam radius (dr/rbeam) on the magnetic field for both cases.

FIG. 11.

Dependencies of the maximum relative Larmor radius on the magnetic field with and without space charge included.

FIG. 11.

Dependencies of the maximum relative Larmor radius on the magnetic field with and without space charge included.

Close modal

The trajectory tracks at the magnetic fields corresponding minimums on curves Fig. 11 are shown in Figs. 12(a) and 12(b).

FIG. 12.

Electron trajectories from the gun in Fig. 9 with (a) and without space charge (b) at magnetic fields set to the minimum value of the trajectory angles and for the electrostatic field corresponding to electron beam generation with current 12.0 A.

FIG. 12.

Electron trajectories from the gun in Fig. 9 with (a) and without space charge (b) at magnetic fields set to the minimum value of the trajectory angles and for the electrostatic field corresponding to electron beam generation with current 12.0 A.

Close modal

These plots demonstrate that with some difference in the shape of curves in Fig. 11 for cases with and without space charge, the general pattern of this dependence seems to be determined primarily by the external fields with the electron space charge playing a secondary role.

Even the distributions of the electron Larmor radius across the beam radius for cases with and without the space charge in Figs. 12(a) and 12(b) show that it is mostly affected by the external fields as well.

The introduction of non-adiabatic fields in the cathode-anode gap of an electron gun has strong effect on the performance of the gun. It changes the dependence of the electron trajectory angle on the magnetic field for a fixed electron current compared to an adiabatic gun. The main benefit of such modification is the possibility to operate the gun at a lower magnetic field than for an adiabatic gun, and for some applications, this benefit is important. The downside of non-adiabatic guns is increased beam oscillations in areas outside the optimum conditions: at different magnetic fields and during the ramping up and down of the electron current. Knowing the effect of field modification, one can optimize the gun to satisfy both the requirements of the minimum magnetic field and acceptable level of oscillations in other important operational regions.

Our simulations show that the observed pattern of the Larmor motion is affected primarily by the external fields with the electron space making much smaller modification to this pattern.

The authors are very grateful to Fredrik Wenander for productive discussions and indispensable help in preparation of this manuscript.

Work was supported by Brookhaven Science Associates, LLC, under Contract No. DE-SC0012704 with the U.S. Department of Energy.

1.
E. D.
Donets
,
V. I.
Ilushchenko
, and
V. A.
Alpert
, JINR (Dubna) P7–4124,
1968
.
2.
V.
Kamerdzhiev
,
G.
Kuznetsov
,
V.
Shiltsev
, and
N.
Solyak
, “
Electron beam generation in tevatron electron lenses
,” FERMILAB-CONF-06-308-AD,
2006
.
3.
W.
Fischer
,
X.
Gu
,
Z.
Altinbas
,
M.
Costanzo
,
J.
Hock
,
C.
Liu
,
Y.
Luo
,
A.
Marusic
,
R.
Michnoff
,
T. A.
Miller
,
A. I.
Pikin
,
V.
Schoefer
,
P.
Thieberger
, and
S. M.
White
, “
Operational head-on beam-beam compensation with electron lenses in the relativistic heavy ion collider
,”
Phys. Rev. Lett.
115
,
264801
(
2015
).
4.
V. E.
Zapevalov
,
V. N.
Manuilov
, and
S. E.
Tsimring
, “
Contribution to the theory of helical beams with trapped electrons
,”
Radiophys. Quantum Electron.
33
(
12
),
1043
1048
(
1990
).
5.
S. E.
Tsimring
, “
Girotron electron beams: Velocity and energy spread and beam instability
,”
Int. J. Infrared Millimeter Waves
22
(
10
),
1433
1468
(
2001
).
6.
D. V.
Kas’yanenko
,
O. L.
Louksha
,
B.
Piosczyk
,
G. G.
Sominsky
, and
M.
Thumm
, “
Low-frequency parasitic space charge oscillations in the helical electron beams of a gyrotron
,”
Radiophys. Quantum Electron.
47
(
5-6
),
414
420
(
2004
).
7.
E. V.
Ilyakov
,
I. S.
Kulagin
,
V. N.
Manuilov
, and
B. Z.
Movshevich
, “
Theoretical and experimental study of the space-charge oscillations in the electron-optical system of a relativistic gyrotron
,”
Plasma Phys. Rep.
37
(
13
),
1139
1144
(
2011
).
8.
M.
Bell
,
J.
Chaney
,
H.
Herr
,
F.
Krienen
,
P.
Møller-Petersen
, and
G.
Petrucci
, “
Electron cooling in ICE at CERN
,”
Nucl. Instrum. Methods Phys. Res.
190
,
237
255
(
1981
).
9.
T. N.
Andreeva
,
I. N.
Meshkov
,
A. N.
Sharapa
, and
A. V.
Shemyakin
, “
The formation of the electron beam with low transverse velocities in systems with an accompanying magnetic field
,” in
13th International Conference on High-energy Accelerators (HEACC 86)
(
Nauka
,
1987
), pp.
351
352
.
10.
A.
Kponou
,
E.
Beebe
,
A.
Pikin
,
G.
Kuznetsov
,
M.
Batazova
, and
M.
Tiunov
, “
Simulation of 10 A electron-beam formation and collection for a high current electron-beam ion source
,”
Rev. Sci. Instrum.
69
(
2
),
1120
1122
(
1998
).
11.
A. N.
Sharapa
,
A. V.
Grudiev
,
D. G.
Myakishev
, and
A. V.
Shemyakin
, “
A high perveance electron gun for the electron cooling
,”
Nucl. Instrum. Methods Phys. Res., Sect. A
406
169
171
(
1998
).
12.
A.
Pikin
,
J. G.
Alessi
,
E. N.
Beebe
,
D.
Raparia
, and
J.
Ritter
, “
Analysis of magneto-immersed electron guns with non-adiabatic fields
,”
Rev. Sci. Instrum.
87
113303
(
2016
).
13.
See www.fieldp.com for the description of used software.