PDF A study of the secondary electron yield /spl gamma/ of ...

[Pages:7]1568

IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 48, NO. 8, AUGUST 2001

A Study of the Secondary Electron Yield

of Insulator Cathodes for Plasma Display Panels

Yasushi Motoyama, Hideomi Matsuzaki, and Hiroshi Murakami

Abstract--In order to provide a guideline in the search for the optimum cathode materials for plasma display panels (PDPs), formulas for the simple calculation of the secondary electron yield were derived from Hagstrum's theory for an insulator without impurity levels. From these, we obtained the generalized relations between and the band parameters of an insulator and the potential energy of an incident particle, which is an ion or a metastable atom. Unlike metals, it is not work function but the sum of band gap and electron affinity that essentially contributes to of an insulator. By applying these formulas, the values of BaO and MgO for He, Ne, Ar, Kr, and Xe ions and metastable atoms were practically calculated. In particular, the metastable atom-induced values of these insulators were calculated for the first time. The values of these insulators for these noble gas ions are determined by Auger neutralization only. As for MgO, which is at present the most useful insulator cathode for PDPs, the values for Kr and Xe ions become zero. These calculated values of MgO for all noble gas ions were compared with experimental results reported previously.

Index Terms--Cathodes, gas discharges, insulators, noble gases, plasma displays, secondary electron emission.

has become more important to study of insulators for various noble gases.

In this paper, formulas for the simple calculation of of an insulator without impurity levels were derived from Hagstrum's theory. From these formulas, we obtained the generalized relations between and the band parameters of an insulator and the potential energy of an incident particle, which is an ion or a metastable atom, to clarify the factors determining . In particular, the metastable atom-induced values of an insulator were calculated for the first time. Moreover by applying these formulas, the values of BaO and MgO for He, Ne, Ar, Kr, and Xe ions and metastable atoms were calculated. These calculated

values of MgO were compared with experimental results reported previously. In general, measurement of for an insulator is not always easy because of possible difficulties due to the charge-up effect and so forth. Therefore, it is useful to clarify the theoretical values of an insulator.

I. INTRODUCTION

A COLOR plasma display panel (PDP) has been rapidly developed [1]?[3] as a wall-mountable high-definition television (HDTV) receiver from such viewpoints as ease of producing a large flat display and high operational speed. However, reductions in the cost and power consumption are required to increase the penetration of PDPs into the marketplace. To reduce power consumption, it is necessary to improve luminous efficiency. One way of doing this is to reduce the discharge voltage. Therefore, the secondary electron yield of the cathode, which closely relates to the discharge voltage, has become an important object of study [4], [5].

As for metal, the generalized relations between values and physical parameters were calculated [6] from Hagstrum's theory [7], [8]. However, as for theoretical values of insulators, except BaO [9], those of MgO for Ne and Ar ions were only calculated [10] after MgO was used as cathodes for monochrome ac PDPs because of its low firing voltage and good durability [11]?[13]. In recent years, color ac PDPs, which use the vacuum ultraviolet radiation from Xe gas to excite phosphors, have been put to practical use. Therefore, it

Manuscript received November 6, 2000; revised February 20, 2001. The review of this paper was arranged by Editor J. Hynecek.

The authors are with the NHK Japan Broadcasting Corporation, Tokyo 1578510, Japan (e-mail: motoyama@strl.nhk.or.jp).

Publisher Item Identifier S 0018-9383(01)05692-1.

II. DERIVATION OF SECONDARY ELECTRON YIELD

A. Conditions for Secondary Electron Emission

According to the research by Hagstrum [7], [8], it is known that secondary electron emission by low-velocity ions, as in conventional gas-discharge phenomena [14], does not depend on kinetic energy but mostly on the potential (internal) energy of the ion. In this case, the mechanism of electron emission consists of the following two processes:

i) Auger neutralization (one step); ii) resonance neutralization Auger deexcitation (two

steps).

Here Auger neutralization, resonance neutralization (the in-

verse of this is resonance ionization), and Auger deexcitation

are tunnel effects as shown in Fig. 1(a)?(c), respectively,

where the latter two are concerned with the excited state of the

atom, especially the metastable states. The notions of physical

parameters used in Fig. 1 are defined in Table I, which are also

used throughout the paper. When impurity levels do not exist

in the band gap of an insulator, the necessary condition that i)

and ii) occur is given by

and

,

respectively. Therefore, depending on the combination of an

insulator and a gas, the following cases exist: only i) or ii) oc-

curs, both i) and ii) occur, or neither i) nor ii) occur. Moreover,

the necessary conditions that the electron can be ejected by

the above mechanisms are

for i) and

for ii).

Here, all electronic transitions between ions (atoms) and a solid

0018?9383/01$10.00 ? 2001 IEEE

MOTOYAMA: SECONDARY ELECTRON YIELD OF INSULATOR CATHODES

1569

Fig. 1. Schematic diagram illustrating electronic transitions at an insulator surface. (a) Auger neutralization of an ion. (b) Resonance neutralization of an ion. (c) Auger deexcitation of on an excited atom.

TABLE I DEFINITION OF PHYSICAL PARAMETERS

electron and is considered proportional to

. Next, for

an electron thus excited to escape from the solid, it is necessary

that

. Assuming that this escape probability is

, we

obtain the following expression for the secondary electron yield

at a distance

(2)

are summarized in Fig. 2, also including the processes without electron emission.

B. Secondary Electron Yield Neutralization

Based on Auger

The electron energy distribution function in the valance

band of the insulator is given by the product of the state density

and the Fermi?Dirac distribution function , where the

latter can be regarded as a step function at room temperature,

that is,

for

and

for

. As

in Fig. 1(a), when electron 1 moves to the ground state of an

atom and electron 2 is excited at the same time, the energy dis-

tribution

of the excited electron is given by the following

expression by using the Auger transform defined in it, with

the assumption that the matrix element of this transition is inde-

pendent of the energies of these electrons [8]

where

for the electron excited by Auger neutralization is

not isotropic. By taking this into consideration, the next formula

was proposed introducing parameters and [8]

(3)

Hagstrum determined

,

by adjusting with

the experimental results of He for Ge. These values are also used

in the present calculation. Since a transition occurs when an ion

comes a long way to the solid surface, the desired must be an

average of over . In practice, however, it is known from ex-

periments that transitions occur collectively at a certain distance

[7], [8]. Therefore, substituting the ionization energy at

for , we can obtain a good approximation of .

In practical calculations, the state density in must

be given. But this varies depending on the kind of insulator. So,

here the calculations are performed for two cases, a flat band

and a parabolic band, to study the influence of the state-density

profile on the value. Furthermore, by putting

and

to normalize the variables, we obtain the following

formula from (2)

(4)

(1) where

where is the delta function of Dirac and shows the conser-

vation of energy, and

is the state density for the excited

; ; .

1570

IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 48, NO. 8, AUGUST 2001

Fig. 2. Chart showing the relations and conditions for various electronic transitions at the solid surface including secondary electron emission.

The function , which is finite only in as follows, depending on the state density assumed:

for flat band

is given Here, as in Auger neutralization, assuming is a flat band

or a parabolic band, and putting

and

, we

obtain

(5)

for parabolic band

(9)

(6)

C. Secondary Electron Yield Based on Auger Deexcitation

When an ion approaches a solid surface and resonance neu-

tralization occurs, the ion becomes the excited atom. After this

process, unless resonance ionization occurs with the condition

for a smaller distance , the excited atom is con-

sidered to return to the ground state by Auger deexcitation. Ac-

cordingly, as a component of the secondary electron yield by

an ion, we have to consider as well as the above-mentioned

. Specifically, when the transition ratio of Auger neutraliza-

tion to resonance neutralization is to

, then is given

by

. On the other hand, in the secondary elec-

tron yield by a metastable atom, this is considered to be

the only component, namely,

.

Similar to the procedure for obtaining , the energy distri-

bution

of the excited electron is given by

(7)

In Auger deexcitation too, by adopting the same escape proba-

bility as in Auger neutralization and the excitation energy

at

, is obtained as follows:

(8)

where

and

. The func-

tion

, which is finite only in

, is given as

follows, depending on the state density assumed:

for flat band (10)

for parabolic band

(11)

III. RESULTS AND DISCUSSION

A. Factors that Determine of Insulator

Formulas (2) and (8) indicate that, unlike metals, it is not work

function but the sum of band gap and electron affinity

that essentially contributes to and of an insulator.

As for , from (4), the necessary condition for

is

, which means

. This is a natural result from

the energy conservation law. Now, defining new quantities

and

and taking the former as a variable and the latter and

as parameters, we calculated for the flat band from (4) and

(5) and that for the parabolic band from (4) and (6), as shown in

Fig. 3(a) and (b), respectively. From these figures, it is evident

that

1) is not more than 0.5 and increases monotonically with increasing ;

2) when is constant, increases with increasing ; 3) when is constant, increases with decreasing ;

MOTOYAMA: SECONDARY ELECTRON YIELD OF INSULATOR CATHODES

1571

0 Fig. 3. Calculated results of the secondary electron yield

based on Auger neutralization as a function of b with the parameters a and : b 2( 1)= 0 0 0 (2  + ) = (E =) 2; a ( )=2 = " =" ; " =" . (a) Flat-band model. (b) Parabolic-band model.

4) the difference in between the flat and parabolic bands increases with decreasing .

From these results, the conditions in searching for an insulator

cathode with large could be as follows: first, should be

small; second, should be small.

On the other hand, as for , from (9), the necessary con-

dition for

is

, which means

. Now,

defining new quantities

and

and taking the former as

a variable and the latter and as parameters, we calculated

for a flat band from (9) and (10), and that for a parabolic band

from (9) and (11), as shown in Fig. 4(a) and (b), respectively.

The situations of depending on , , , and the state-den-

sity profile are similar to those of depending on , , and

the state-density profile. Also, the conditions in searching for an

insulator cathode with large are similar to those of .

and become the maximum limiting values when the

electron affinity becomes zero (namely

) as shown in

Figs. 3 and 4. It may be considered that these limiting cases also

hold for the insulators with negative electron affinity (NEA).

B. Calculation of Values of BaO and MgO

In the practical calculations of , the values of and are necessary, which are not always obvious. However, it was reported that the difference between and (free space ionization energy) was small, for example, within about 10% [8]. Therefore, for simplicity, calculations were made on the assumption that the values of and were equal to those of

and (free space energy of the metastable state), respectively.

The values of BaO and MgO for all noble gas ions and

metastable atoms were calculated by using the formulas in the

previous section. Table III shows these results. The band param-

eters of BaO and MgO used in these calculations are shown in

Table II. The calculated results show that resonance neutraliza-

tion can not occur with the combinations of these insulators and

all noble gases because of the condition

. There-

fore, values of BaO and MgO for all noble gas ions are deter-

mined by Auger neutralization only, namely

.

The values of BaO for all noble gas ions are similar to those

reported in [9]. As for MgO, the calculated values for Kr

and Xe ions become zero because of the condition

,

although values remain finite for all noble-gas metastable

atoms. This result is remarkable, since the Xe ions are consid-

ered principal incident particles toward cathodes for practical

color PDPs.

C. Comparisons with Experimental Values of MgO

In general, measurement of for an insulator is not always easy because of possible difficulties due to the charge-up effect and so forth. Therefore, as shown in Table IV, measured values of MgO do not always show a constant value in these experiments done by different researchers. In three cases, the values were measured using an ion beam technique in a vacuum [15], [17], [18], in which case the measured values depend on the accelerating voltage of the ions. In the other cases, values were estimated from breakdown voltages in gas [16], [19], [20], in which case the estimated values depend on ( is the electric field, is the gas pressure). Therefore, it is difficult to compare our results with such experimental results quantitatively.

1572

IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 48, NO. 8, AUGUST 2001

0 Fig. 4. Calculated results of the secondary electron yield

based on Auger deexcitation as a function of b with the parameters a and : b ( 1)= 0 0 0 (1  + ) = (E =) 1; a  = " =" ; " =" . (a) Flat-band model. (b) Parabolic-band model.

TABLE II BAND PARAMETERS OF BaO AND MgO USED IN THIS STUDY

TABLE III

E E

VALUES OF AND

FOR NOBLE GASES AND THE THEORETICALLY CALCULATED VALUES OF BaO AND MgO

However, it is useful to compare the general tendency of our results with those of the experimental results in Table IV. Here, the variation tendency of the experimental values depending on the kind of gas ions, coincides well with that of our results. Especially, the small values for the Xe ion in these experiments support the validity of our model.

D. Dependence of Values of MgO on State-Density Profile

Figs. 5 and 6 show the values of MgO versus and the values of MgO versus respectively. The values for

the flat band become larger than those for the parabolic band with decreasing . The values for the flat band also become larger than those of the parabolic band with decreasing

. These result from the fact that when and are small, electron transitions principally occur in the upper level of the valence band, where the state density of the flat band is larger than that of the parabolic band. As for and values for noble gases, however, the differences between those for flat and parabolic bands are small as shown in Figs. 5 and 6. Therefore, it can be considered that and of MgO for the noble gases are almost independent of their state-density profile.

MOTOYAMA: SECONDARY ELECTRON YIELD OF INSULATOR CATHODES TABLE IV

COMPARISONS OF THE THEORETICALLY CALCULATED VALUES OF MgO WITH EXPERIMENTAL VALUES FROM THE LITERATURE

1573

Fig. 5. Calculated

values of MgO as a function of b both for the flat-band

0 model (broken line) and parabolic-band model (solid line): b 2( 1)= 0 0 (2  + ); a ( )=2 = " =" = 0:395; " =" = 0:933.

neutralization. As for MgO, the values for Kr and Xe ions become zero, although values remain finite for all noble-gas metastable atoms. The calculated values of MgO for all noble gas ions were compared with experimental results reported previously. In general, the measurement of of insulators is not always easy because of possible difficulties due to the charge-up effect and so forth. Therefore, it is useful to clarify the theoretical values of insulators without impurity levels as a first step. These results should act as a useful guideline in the search for the optimum cathode materials for a PDP.

As for further work, a study into the influences of

1) charge-up effect; 2) impurity levels; 3) crystalline orientation; 4) adsorption etc., on of insulators is needed.

Furthermore, in the PDP-like gas space, due to the back diffusion of electrons, effective secondary electron yield was known to be reduced in comparison with the present in a vacuum [21]. This is also being studied in our research group [22].

Incidentally, after our report on the theoretical study of the of insulators [23], [24], we found a report on a theoretical study dealing with the particular case of MgO [25]. Although this report does not contradict ours, the value of MgO calculated in this report is larger than ours, probably due to the different formula used for escape probability.

ACKNOWLEDGMENT

The authors would like to thank Dr. F. Sato and all members of the research group for their encouragement and useful suggestions.

Fig. 6. Calculated

values of MgO as a function of b both for

the flat-band model (broken line) and parabolic-band model (solid line):

0 0 0 0 b ( 1)=(1  + ) = (E =) 1; a  = " = " = 0:395; " =" = 0:933.

IV. CONCLUSION

We obtained the generalized relations between and the band parameters of an insulator and the potential energy of an incident particle, which is an ion or a metastable atom. Moreover, the values of BaO and MgO for He, Ne, Ar, Kr, and Xe ions and metastable atoms were calculated. The results show that the

values for these noble gas ions are determined only by Auger

REFERENCES

[1] T. Yamamoto, T. Kuriyama, M. Seki, T. Katoh, T. Takei, T. Kawai, H. Murakami, and K. Shimada, "40-inch-diagonal HDTV DC plasma display," IEEE Trans. Electron Devices, vol. 42, pp. 847?855, May 1995.

[2] M. Seki, Y. Takano, T. Katoh, T. Yamamoto, T. Kawai, T. Koura, S. Ueda, T. Takei, T. Tajima, T. Kuriyama, J. Koike, H. Murakami, K. Takahashi, Y. Sasaoka, and M. Kasahara, "Development of a 42-inch DC-PDP for Hi-vision" (in Japanese), J. Inst. Image Inform. Television Eng., vol. 54, no. 2, pp. 301?309, 2000.

[3] T. Nishio and K. Amemiya, "High-luminance and high-definition 50-in.diagonal co-planar color PDPs with T-shaped electrodes," in Proc. SID Int. Symp. Dig. Tech. Papers, May 1999, pp. 268?271.

[4] O. Sahni and C. Lanza, "Importance of the dependence of the secondary

electron emission coefficient on E=p for Paschen breakdown curves in

ac plasma panels," J. Appl. Phys., vol. 47, no. 4, pp. 1337?1340, 1976. [5] M. Ishimoto, D. Hidaka, K. Betsui, and T. Shinoda, "Secondary-electron

analysis of MgO films in ac plasma displays," in Proc. SID Int. Symp. Dig. Tech. Papers, May 1999, pp. 552?555.

1574

IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 48, NO. 8, AUGUST 2001

[6] H. Matsuzaki, "Discharge characteristics and their analysis of alka-

line?metal cold cathodes" (in Japanese), Trans. Inst. Elect. Eng. Jpn.,

vol. 111-A, no. 11, pp. 971?978, 1991.

[7] H. D. Hagstrum, "Theory of Auger ejection of electrons from metals by

ions," Phys. Rev., vol. 96, no. 2, pp. 336?365, 1954.

[8]

, "Theory of Auger neutralization of ions at the surface of a di-

amond-type semiconductor," Phys. Rev., vol. 122, no. 1, pp. 83?113,

1961.

[9] Y. Takeishi, "Ejection of electrons from barium oxide by noble gas ions,"

J. Phys. Soc. Jpn., vol. 17, no. 2, pp. 326?341, 1962.

[10] M. O. Aboelfotoh and J. A. Lorenzen, "Influence of secondary-electron

emission from MgO surfaces on voltage-breakdown curves in Penning

mixtures for insulated-electrode discharge," J. Appl. Phys., vol. 48, no.

11, pp. 4754?4759, 1977.

[11] T. Urade, T. Iemori, M. Osawa, N. Nakayama, and I. Morita, "A pro-

tecting layer for the dielectric in ac plasma panels," IEEE Trans. Elec-

tron Devices, vol. ED-23, pp. 313?318, Mar. 1976.

[12] H. Uchiike, K. Miura, N. Nakayama, T. Shinoda, and Y. Fukushima,

"Secondary electron emission characteristics of dielectric materials in

ac-operated plasma display panels," IEEE Trans. Electron Devices, vol.

ED-23, pp. 1211?1217, Nov. 1976.

[13] N. J. Chou and O. Sahni, "Comments on `Secondary electron emis-

sion characteristics of dielectric materials in ac-operated plasma dis-

play panels'," IEEE Trans. Electron Devices, vol. ED-25, pp. 60?62,

Jan. 1978.

[14] Y. K. Shin, J. K. Lee, C. H. Shon, and W. Kim, "Ion energy distribution

in alternating-current plasma display panel cell," Jpn. J. Appl. Phys., pt.

2, vol. 38, no. 2B, pp. L174?L177, 1999.

[15] N. J. Chou, "Ion-induced secondary-electron emission from MgO

films," J. Vac. Sci. Technol., vol. 14, no. 1, pp. 307?311, 1977.

[16] S. Ho, M. Saji, S. Ihara, M. Shiiki, K. Suzuki, A. Yuhara, A. Yokoyama,

M. Ishigaki, R. Sato, N. Kouchi, and Y. Hatano, "Numerical analysis of

discharge voltage and light emission efficiency in ac-PDPs," in Proc.

Int. Display Workshops'98, Dec. 1998, pp. 479?482.

[17] K. S. Moon, J. Lee, and K.-W. Whang, "Electron ejection from MgO

thin films by low energy noble gas ions: Energy dependence and ini-

tial instability of the secondary electron emission coefficient," J. Appl.

Phys., vol. 86, no. 7, pp. 4049?4051, 1999.

[18] E.-H. Choi, J.-Y. Lim, Y.-G. Kim, J.-J. Ko, D.-I. Kim, C.-W. Lee, and

G.-S. Cho, "Secondary electron emission coefficient of a MgO single

crystal," J. Appl. Phys., vol. 86, no. 11, pp. 6525?6527, 1999.

[19] G. Auday, P. Guillot, and J. Galy, "Secondary emission of dielectrics

used in plasma display panels," J. Appl. Phys., vol. 88, no. 8, pp.

4871?4874, 2000.

[20] V. V. Elsbergen, P. K. Bachmann, and T. Juestel, "Ion-induced secondary

electron emission: A comparative study," in SID Int. Symp. Dig. Tech.

Papers, May 2000, pp. 220?223.

[21] O. Sahni and C. Lanza, "Influence of the secondary electron emission

+ coefficient of argon on Paschen breakdown curves in ac plasma panels

for neon 0.1% argon mixture," J. Appl. Phys., vol. 47, no. 11, pp.

5107?5108, 1976.

[22] Y. Murakami, H. Matsuzaki, H. Murakami, and N. Ikuta, "Effective sec-

ondary electron yield of a cathode for plasma display panel," Jpn. J.

Appl. Phys., no. 5A, pp. 3382?3388, 2001.

[23] Y. Motoyama, M. Ushirozawa, and H. Matsuzaki, "Calculation of the

secondary electron yield

of insulator for PDP cathode," (in Japanese),

IEICE Tech. Rep. EID99-46, July 1999.

[24] Y. Motoyama, M. Ushirozawa, H. Matsuzaki, Y. Takano, and M. Seki,

"Study on the secondary electron yield

of insulator for PDP cathode,"

Bull. Am. Phys. Soc. (52 Gas. Electron. Conf.), vol. 44, no. 4, p. 52, Oct.

1999.

[25] J. Yoon, I. Lee, J.-W. Lee, and B. Oh, "Intrinsic and effective secondary electron emission coefficients in ac plasma display panel," Jpn. J. Appl. Phys., vol. 2A, pp. 809?812, 2001.

Yasushi Motoyama received the B.E. and M.E. degrees in electrical information engineering from Yokohama National University, Japan in 1984 and 1986, respectively.

In 1986, he joined the NHK Japan Broadcasting Corporation. From 1986 to 1989, he worked at the NHK Niigata Broadcasting Station, Niigata Prefecture. Since then, he has been with the Science and Technical Research Laboratories, NHK, Tokyo, and has been engaged in research of plasma display devices. Mr. Motoyama is a member of the Institute of Image Information and Television Engineers.

Hideomi Matsuzaki was born in Tokyo, Japan on April 28, 1944. He received the B.E. degree in applied physics, and the M.E. and Ph.D. degrees in electronic engineering, all from the University of Tokyo, Japan, in 1967, 1969, and 1972, respectively.

Since 1972, he has been with the Science and Technical Research Laboratories, NHK Japan Broadcasting Corporation, Tokyo, where he has been engaged in the research and development of plasma display panels. His current research interest is electrical and optical phenomena in gases and their applications. Dr. Matsuzaki is a member of the Institute of Image Information and Television Engineers and the Institute of Electrical Engineers of Japan.

Hiroshi Murakami received the B.E. degree in electrical communication engineering and the Ph.D. degree in electronic engineering from Osaka University, Japan, in 1967 and 1988, respectively.

In 1967, he joined the NHK Japan Broadcasting Corporation. He has been engaged in the research and development of plasma displays at the Science and Technical Research Laboratories, NHK, Tokyo, since 1967, except one year from 1998 to 1999, when he was the Secretary-General of the Hi-Vision PDP Consortium, NHK Engineering Services, Inc. Dr. Murakami received the SID Special Recognition Award in 1994 and awards for his contribution to the development and commercialization of HDTV plasma displays from both the ITE and the IEICE in 2000. He is a member of the Society for Information Displays, the Institute of Image Information and Television Engineers, and the Institute of Electronics, Information, and Communication Engineers.

 ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download