ISSN 10637842, Technical Physics, 2012, Vol. 57, No. 8, pp. 1101–1105. © Pleiades Publishing, Ltd., 2012. Original Russian Text © V.A. Burdovitsin, A.K. Goreev, A.S. Klimov, A.A. Zenin, E.M. Oks, 2012, published in Zhurnal Tekhnicheskoi Fiziki, 2012, Vol. 82, No. 8, pp. 62–66.

GAS DISCHARGES, PLASMA

Expansion of the Working Range of Forevacuum Plasma Electron Sources toward Higher Pressures V. A. Burdovitsin*, A. K. Goreev, A. S. Klimov, A. A. Zenin, and E. M. Oks Tomsk State University of Control Systems and Radioelectronics, pr. Lenina 40, Tomsk, 634050 Russia *email: [email protected] Received October 11, 2011

Abstract—It is shown that the pressure in forevacuum plasma electron sources is limited from above by a cur rent component that arises in the accelerating gap from a highvoltage glow discharge and dominates in the electron beam. The working pressure range of such electron sources can be expanded toward higher pressures by limiting the current of the highvoltage glow discharge in the accelerating gap. DOI: 10.1134/S1063784212080075

INTRODUCTION Generation of stationary and quasistationary elec tron beams under elevated gas pressures with subse quent extraction of the electron beam to the atmo sphere is a priority area of electron beam technology [1]. Plasmacathode electron sources are today viewed as the most promising devices for this purpose [2, 3]. They emit electrons from the surface of a plasma gen erated by a system with a coldcathode glow or arc dis charge (by the cold cathode, it is meant a cathode that is not heated to thermionic temperatures). These sys tems offer a number of fundamental advantages, of which insensitivity to the residual gas atmosphere is the most important. This is why the use of plasma elec tron sources is viewed as an effective way of expanding the working range of electron beam generators toward higher pressures. Forevacuum plasma sources, which have been extensively developed in recent years, generate both continuous and pulsed beams with various configura tions in the higher than ever forevacuum pressure range up to 10–15 Pa [4] (by the forevacuum pressure range, here it is meant the pressure range provided by one stage of a mechanical pump.) Such an advance toward higher working pressures has enhanced the capabilities of electronbeam modification of materi als. Specifically, there has appeared the possibility of electronbeam processing of nonconductive ceramics [5]. Further advance toward still higher pressures is of interest for further progress in electron beam genera tion methods and for their wider application. Obvi ously, the pressure at which an electron beam is gener ated is limited by the initiation of a lowvoltage self maintained discharge in the accelerating gap. In this case, electrons cannot be accelerated to a desired energy. Such conditions are attained when the product pd reaches a minimum in the Paschen curve. Although in forevacuum plasma sources of electrons, the oper

ating point pd is still in the lefthand branch of the Paschen curve, the dielectric strength failure of the accelerating gap still remains a main problem adversely affecting the efficiency of such devices. The stability of such electron sources depends not only on the electron beam but also on a “parasitic” highvolt age glow discharge (HVGD) arising in the accelerating gap. Despite physical reasons limiting the working pressure of forevacuum plasma electron sources being obvious, their limit pressure is still poorly understood. In this work, we try to elucidate this question. EXPERIMENT Experiments were carried out with a plasma elec tron source (Fig. 1) based on a hollowcathode glow discharge initiated to generate a beam in the forevac uum pressure range [6]. The accelerating gap of the electron source was configured so as to prevent break down along “long paths” and thereby generate a sta tionary focused electron beam at elevated pressures. An emitting plasma arises in a discharge ignited by 55mmlong hollow cathode 1 20 mm in diameter. Plane anode 2 has emission window 3 10 mm in diam eter. The emission window is shut by a thin (~1 mm) tantalum or molybdenum foil with perforated holes 0.7 mm in diameter. An accelerating gap is formed by an emitting electrode (the outer surface of anode 2) and grounded accelerating electrode 4. Electron beam 5 focused by magnetic system 6 falls on collector 7. All electrodes of the electron source, except the perfo rated plate, are made from stainless steel. The electron source operated in the continuous mode with the fol lowing parameters: a discharge current of 0.05– 0.30 A, a beam current up to 0.2 A, an accelerating voltage of 5–20 kV, and a beam diameter on the col lector of 5–10 mm.

1101

1102

BURDOVITSIN et al. 1

Id

Ud

3

200

Ie

Ua

Ie, mA

2 4

(a)

1 2 3 4 5

250

150 100 50

6

0 5

0

7

Ie, mA

10

15

1 2 3 4 5

120 100

Fig. 1. Plasma source of electrons: (1) hollow cathode, (2) anode, (3) emission window, (4) accelerating electrode, (5) electron beam, (6) focusing system, and (7) collector.

5

80

(b)

60 40

In the forevacuum pressure range, it is impossible to provide a pressure drop between the discharge and accelerating gaps (as is usually done in plasma electron sources designed for the pressure range 0.01–0.10 Pa, which is typical of such devices) [7]. The isobaric operating mode of forevacuum plasma sources of elec trons is another basic feature of these devices, which directly influences the dielectric strength of the accel erating gap. In experiments, air and helium were used as working gases. Air was applied largely for modeling the practical application of forevacuum plasma source of electrons, while helium was used as a gas convenient for expanding the pressure range. The pressure in the electron source was varied from several pascals to sev eral tens of pascals by letting the gas flow directly into the vacuum chamber. The experiments were aimed at achieving the ulti mate working pressure of the forevacuum plasma source of electrons. They were conducted as follows. The working chamber was filled with a gas to a certain pressure, and a discharge generating an emitting plasma was initiated in the plasma source of electrons between hollow cathode 1 and anode 2. Then, voltage Ua across the accelerating gap was raised and an elec tron beam was formed. After voltage Ua reached some critical value, the accelerating gap experienced break down. Breakdown was detected from a sharp drop of the voltage and a rise in current Ie in the load circuit of a highvoltage rectifier feeding the accelerating gap. The instant of breakdown was fixed from the disap pearance of the electron beam on the collector. In a number of experiments, the discharge voltage was not applied on electrodes 1 and 2. In this case, the dielec

20 0 0

5

10 Ua, kV

15

20

Fig. 2. I–U characteristics of the forevacuum plasma elec tron source at a discharge current of (1) 0, (2) 100, (3) 200, (4) 250, and (5) 300 mA in (a) air and (b) helium. p = 10 Pa.

tric strength of the accelerating gap was determined in the “cold” mode (without the electron beam). RESULTS AND DISCUSSION The typical current–voltage (I–U) characteristics of the electron source in air and helium are presented in Figs. 2a and 2b, respectively. Remarkably, even in the absence of the discharge current and, conse quently, the emission current from the plasma, an electron beam with an appreciable current arises in response to a highvoltage applied to the accelerating gap. Naturally, a hollowcathode discharge ignited in the electrode system of the forevacuum plasma source of electrons increases the beam current in the gap but lowers the limiting accelerating voltage at which the electron source is still stable. The greater the discharge current, the lower the accelerating voltage causing the gap to break down (Fig. 2a). The rise in the dielectric strength and operation stability of the forevacuum plasma source of electrons observed when air was TECHNICAL PHYSICS

Vol. 57

No. 8

2012

EXPANSION OF THE WORKING RANGE

replaced by helium (Fig. 2b) was expectable. In addi tion, the increase in the gas pressure limits the ulti mate voltage across the accelerating gap and, hence, the maximal energy of the electron beam (Fig. 3). In this case also, helium is preferable as a working gas. It is seen from Fig. 3 that helium provides stable genera tion of an electron beam at higher pressures up to 30 Pa. The appearance of the electron beam in the accel erating gap in the absence of plasmaemitted electrons seems to be associated with an HVGD arising in the gap. In this case, the true emission current is added to the electron current of the HVGD, producing a total current of the electron beam. Since HVGD electrons are knocked out of the outer surface of the anode and then accelerated by the total applied voltage, they are indistinguishable from the plasmaemitted electrons. When the pressure rises, the HVGD current in the gap increases considerably (Fig. 3) and, accordingly, so does the electron component of the HVGD in the cur rent of the accelerated electron beam. Under certain conditions, the HVGD electron current begins to exceed the plasmaemitted electron current. Although a high voltage persists in the accelerating gap in this case too and an accelerated electron beam forms, such an operating mode of the plasma source of electrons should be considered as inefficient. The fact is that the dominance of the HVGD component in the electron beam prevents energyindependent control of the beam current, which in this case depends mostly on the conditions in the accelerating gap, rather than on the parameters of the emitting plasma. Thus, the pressure range of the forevacuum plasma source of electrons is basically limited by the domi nance of the HVGD current in the total current of the accelerated electron beam. Such a situation may arise in the pressure range where the accelerating gap still retains a dielectric strength and an electron beam forms. Since the transition to the HVGD with rising operating pressure is smooth, the boundary between electron beam generation via electron emission from the plasma and via ion–electron emission from the surface of the emitting electrode in the course of HVGD development is smeared. Apparently, categori zation of the electron beam generation modes by prev alence of the plasma or ion–plasma component in the beam is more appropriate. The analysis of the electron source at the instant preceding the breakdown of the accelerating gap did not reveal any essential difference between the cases with and without the discharge current, which is responsible for the plasma emission component in the beam. At each instant of breakdown, the current in the feeding circuit of the accelerating gap changed step wise without affecting the discharge gap. This suggests that the breakdown of the accelerating gap in the for evacuum plasma source of electrons is due to the beam plasma generated in the electron beam transport region adjacent to the accelerating electrode. Unlike the conditions considered in [9], where breakdown TECHNICAL PHYSICS

Vol. 57

No. 8

2012

1103

Ie, mA 1 2 3 4

150

100

50 0

5

10

15

20 Ua, kV

Fig. 3. I–U characteristics taken at a pressure of (1) 6, (2) 10, (3) 20, and (4) 30 Pa in helium. Id = 200 mA.

was observed when the plasma escaped through the emission hole into the accelerating gap, we are dealing with breakdown between the beam plasma and the negatively biased emission electrode of the accelerat ing gap. This, in turn, means that processes taking place on the surface of this electrode are of primary importance in breakdown initiation. It is natural to suppose that, as the gas pressure and the emission cur rent grow, so does electric field strength E near the sur face of this electrode, since the beam plasma concen tration rises and the plasma boundary approaches the emitting electrode. As a result, emission centers arise with subsequent development of cathode spots and arcing. To make quantitative estimates, we will proceed as follows. Concentration n of the beam plasma is found from the ion balance equation 2kT e je D n = σn 0  , Mi q

(1)

where Te is the electron temperature of the plasma, Mi is the ion mass, σ is the cross section of gas ionization by beam electrons, je is the beam current density, n0 is the concentration of neutral gas molecules, and D is the characteristic longitudinal size of the plasma that is equal to its diameter and at which a 1D approxima tion is still applicable. In the forevacuum pressure range, je consists of two components, j e = αU a + j pl ,

(2)

where the first term is related to the HVGD and the second one is related to the electron emission from the discharge plasma. Coefficient α taking into account the dependence of current je on accelerating voltage Ua

1104

BURDOVITSIN et al.

Ub, kV

Ie, mA 200

20 1 2 3 4

15

10

1 2 3 4

150

100

50 5 0 0 0

50

100

150 jpl, mA/cm2

0

5

10

15

20 Ua, kV

Fig. 4. Breakdown voltage Ub vs. plasma emission current density at a pressure of (1, 2) 10 and (3, 4) 20 Pa. (1, 3) Experiments and (2, 4) calculation.

Fig. 5. I–U characteristics (1, 3) before and (2, 4) after the diaphragm was placed in the accelerating gap. Air pressure p = (1, 2) 6 and (3, 4) 20 Pa.

was found experimentally. Under the assumption of like charges in a layer between the plasma electrodes and the emitting electrode, thickness d of this layer can be found from the wellknown expression [9]

nected to the emitting electrode. As follows from Fig. 5, this considerably decreases the HVGD current and raises the ultimate pressure. Simultaneously, the growth of the ultimate accelerating voltage was observed.

1/2

3/4

ε0 Ua (3) d =  , 1/2 1/4 n ( qkT e ) where Ua is the potential drop across the layer, which equals the accelerating voltage in the given case. Taking into account that d = Ua/E, expressing n from (1), and substituting the resulting expression into (3), we obtain on rearrangements ⎛ ε 0 q 1/2 E 2 ⎞ 1 αU a + j pl = ⎜  ⎟ . ⎝ σn 0 DM 1/2⎠ U 1/2 a

(4)

Equation (4) can be used to analyze the depen dence of breakdown voltage Ub on the pressure and plasma emission current if critical field strength Ecr at which emission centers arise on the emitting electrode is known. Clearly, the value of Ecr can be set only ten tatively; nevertheless, at E = Ecr = 2 × 104 V/cm, the above dependence agrees well with experimental data (Fig. 4). This is an additional argument in favor of the breakdown mechanism suggested. It is reasonable to suppose that the working pres sure of the forevacuum plasma source of electrons can be raised further by providing conditions reducing the field strength at the emitting electrode. This would make it possible to increase critical value Ecr of the mean electric field and, consequently, ultimate accel erating voltage Ub. In this work, the pressure was increased by placing a diaphragm between the acceler ating and emitting electrodes that is electrically con

CONCLUSIONS Thus, when the working pressure of a forevacuum plasma electron source increases, a considerable frac tion of the electron component of a highvoltage glow discharge initiated in the accelerating gap appears in the electron beam. Above a certain pressure, the elec tron current of the glow discharge dominates in the total current of the accelerated beam. Although the accelerating gap remains electrically strong and an electron beam can be formed under such conditions, this situation should be considered as marginal for the effective operation of forevacuum plasma source of electrons. The working pressure range can be expanded toward higher values if conditions limiting the highvoltage glow discharge current in the acceler ating gap are provided. ACKNOWLEDGMENTS This work was supported by the Russian Founda tion for Basic Research, grant. no. 110800074a. REFERENCES 1. A. Hershcovitch, J. Appl. Phys. 78, 5283 (1995). TECHNICAL PHYSICS

Vol. 57

No. 8

2012

EXPANSION OF THE WORKING RANGE 2. Yu. E. Kreindel’, Plasma Electron Sources (Atomizdat, Moscow, 1977). 3. E. M. Oks, Plasma Cathode Electron Sources: Physics, Technology, Applications (NTL, Moscow, 2005; Wiley, New York, 2007). 4. V. A. Burdovitsin and E. M. Oks, Laser Part. Beams 26, 619 (2008). 5. V. A. Burdovitsin, A. S. Klimov, and E. M. Oks, Tech. Phys. Lett. 35, 511 (2009).

TECHNICAL PHYSICS

Vol. 57

No. 8

2012

1105

6. V. A. Burdovitsin, I. S. Zhirkov, E. M. Oks, I. V. Osi pov, and M. V. Fedorov, Prib. Tekh. Eksp., No. 6, 66 (2005). 7. I. Osipov and N. Rempe, Rev. Sci. Instrum. 71, 1 (2000). 8. Yu. E. Kreindel’ and V. A. Nikitinskii, Tech. Phys. 26 (11), (1971). 9. M.A. Zav’yalov, Yu. E. Kreindel’, A. A. Novikov, and L. P. Shanturin, Plasma Processes in Technological Elec tron Guns (Energoatomizdat, Moscow, 1989).