Sputtered thermionic hexaboride coatings

SURfACE

&COATINIiS

ELSEVIER

Surface and Coatings Technology 98 (1998) 1315-1323

HCHNOJDGY

Sputtered thermionic hexaboride coatings W. Waldhauser a .• , C. Mitterer a, J. Laimer b, H. Stori b a

Institutjiir Metallkunde und Werkst()!fpriifunK. Montanunil'ersitiit. Fran:-Josej-StrajJe 18, A-8700 Leohen. Austria b Institutjar AIIKemeille Physik, Tecllllische Ullit'ersitiit. WiednerllauptstrajJe 8-10, A-UJ40 Wien. Austria

Abstract Thermionic coatings based on the hexaborides of rare-earth elements (ReB 6 ) were deposited onto molybdenum and tungsten substrates using DC magnetron sputtering from LaB 6 • CeB 6 , 5mB6 and YB 6 targets. Films were investigated by means of scanning electron microscopy, electron probe microanalysis and X-ray diffraction. The work function and electron emission characteristics of the coatings have been studied by the thermionic emission method using coated tungsten filaments. After optimization of the sputtering parameters. extremely fine-columnar coatings consisting of predominantly (100)-orientated LaB 6 • CeB 6 and 5mB6 crystals were obtained. The Y-B films showed nearly amorphous film growth. All coatings were overstochiometric with B:Re atomic ratios ranging from 6.3 to 7.5. The work function was measured to be in the range 2.6-3.3 eV. Coated cathodes worked with comparable electron emission current at temperatures approximately 1000"C below the operating temperature of uncoated tungsten filaments. LaB 6 and CeB6 were found to be the most promising electron emissive coating materials useful as low temperature emitters. However, satisfactory lifetimes may be only yielded at low operating temperatures and vacuum pressures. f) 1998 Elsevier Science S.A. Keyll'Ords: Electron emission; Magnetron sputtering; Physical vapour deposition; Rare-earth hexaborides; Thermionic coatings

1. Introduction The rare-earth metals (Re) form six different families of binary borides: ReB 2, ReB 4 • Re2BS, ReB 6 , ReB 12 and ReB,., 66 [1,2]. The structure of these borides is built from different boron basic units: sp2-type (two-dimen• sional network in hexagonal diborides); B6 -type (octahe• drons in cubic hexaborides); and B 12 (e,g. cubo• octahedrons in cubic dodecaborides). All boron sublat• tices are electron-deficient and require electron transfer from the metal atoms in order to be stabilized. A rough examination of the rare-earth metals able to form borides shows that their atomic size rather than their electron configuration is the principal factor governing the boride structure [3]. As a result of their specific crystal chemistry, the hexaborides of the rare-earth elements are interesting materials for cathode applica• tions since they combine properties such as low work functions, good electrical conductivities and high melting points, as well as low volatility at high temperatures [4• 7]. At present, either rods of sintered LaB 6 or single LaB 6 crystals are indirectly heated to induce emission • Corresponding author. 0257-8972/98/$19.00 © 1998 Elsevier Science S.A. All rights reserved. PI/ S0257·8972(97)00263-6

or metal filaments such as tungsten, rhenium or tantalum are electrophoretically coated with LaB 6 , sintered and directly heated. Representing very good emitters, both types of cathodes are generally fragile and can withstand only a limited number of thermal shocks. In addition, indirectly heated LaB 6 electron sources complicate the design of electron guns and consume a higher amount of heating power. Single crystals of LaB 6 are rather expensive. Electrophoretically coated metal filaments being simple in design were described by Favreau [8] and Khairnar et al. [9]. Unfortunately, these deposition methods produce porous coatings with low adhesion [10]. In general, coatings fabricated by PVD techniques are more dense, adherent and smoother than coatings deposited by electrophoretic methods. Mroczkowski [10] deposited LaB 6 onto tungsten and rhenium filaments employing RF magnetron sputtering and obtained coat• ings with work function values of 2.4-2.6 eV. Yutani et al. [II] prepared thin LaB 6 films at several substrate temperatures by electron beam evaporation with different angles of incidence of the particle flux and yielded work function values of 2.7-2.8 eV. Okamoto et a1. fabricated plasma display devices with LaB 6 -coated

nickel cathodcs, which shO\led Illlrk I'unctions 01' 2.4 2.5 eV [121. Nakal1l1 et al. produced l.aB" Iilm~ by RF sputtering and discussed the rclationshirs hctween internal stress and the orientation Ill' Ii 1m crystallites [131. Only a lery limitcd number 01' rarers have hcen puhlished on the depositillll and electron emission 01' other rare-earth he\ahondc tilms such as YB" [1.+.15J or 5mB,> [Ib]. In this parer. lIe report on structure. morphology. thermal stahility and electron emission characteristics 01' Re B l'Oatings (where Re = La. Ce. Sm. Y) deposited hy non-reactive DC magnetron sputtering using ReB" targets.

2. Experimental The coatings were I'ahricated hy non-reactive DC mag• sputti:ring in a modified sputtering unit (Consolidated Vael.:um Corporation cve AST 60 I USA) using commercially availahle LaB". CeB". 5mB" and YB" targets with a diameter of75 mm and a thickness 1ll.~tron

01' (, mm. The suhstrates used were ground molybdenum and metallographically polished tungsten sheets with thicknesses or 0.5 mm 1'01' structural investigations as well as various Cllll1mlTcially available tungsten cathodes (filament diameter 0.1 mm) I'or thermionic emission measurements. All suhstrates were precleaned with ethy• lene in an ultrasonic cleaner and could he either grounded or biased up to - 1600 V Cor sputter cleaning prior to deposition. Deposition was carried out at a substrate temperature or ahout 200 C in the static (suhstrates without movement) and dynamic mode (rotating suh• strates) using argon at pressures ranging Crom I to 4 Pa. The target substrate distance Ivas 60 mm ror static deposition. The thil.:knesses or those l'Oatings deposited onto molyhdenum and tungsten sheets reached 3 ± I ~lm. The dynamic deposition was carried out onto rotating suhstrates at a mean target substrate distance or b2 mm and ri:sultcd in coating thicknesses 01'2 ±O.5 pm. Coating morphology and structure were investigated hy means or scanning electron microscopy (SEM ) and X-ray di/frac• tion ( XRD) using a Cambridge Instruments Stercoscan 360 scanning electron microscope (England) and a .

[l

[HI- 25.0 KV

UO- 12

...

[HT .. 25 0 lV

olD- 14

rTIf'I

.oo""~

t-

[;

[111.

'!:j

loln·

'J (J

'il,Jm~-

~

-

11 --1

!Ml

PHOTO· 136,04

L" S£1

Z OOflmf

PHOIO- 6736

PHOrn.

~

',,(

---~

hg. I. ('rllSs-s~clil)1lal SLM micrographs 01" sputt~r~d rare-earth he.xaooride coatings. Static Ikposition: suostrak. molyndenum: spullenng power density. 4.S W crn ': sunstrak t~mperature. 21l1l C tal L.aB,,-nased coating (argon pr~ssure. :I Pa: nias voltage. IIlIl V): (b I C~B,,-nased coating (argon pressure,:I I'a: grounded sunstrate): (1'1 5mB,,-nased coating (argon pressure. J I'a: nias voltage. -11l1l V); and (d) YB,,-nased coating (argon pr~ssure. 2 I'a: nias voltage. - IIlIl V I.

W. Waldhauser et al.

I Surface and Coatings Teclmology 98 (1998) 1315-1323

diode test cell using coated tungsten filaments [17]. Application-related properties such as current stability and lifetime were investigated by testing coated tungsten hairpin cathodes in the above described SEM and in the diode test cell.

Siemens F goniometer (Germany) with CrKoc radiation. The chemical composition of the deposits was determined by electron probe microanalysis (EPMA) employing a Microspec WDX-3PC system (USA) mounted to the above mentioned SEM. Quantification occurred by means of commercially available LaB 6 , B, Y, CeF 3 and SmF 3 standards. Heat treatment up to 900°C in a resis• tively heated furnace in argon atmosphere was carried out to investigate the thermal stability of the coatings deposited onto tungsten sheets. The electron emission measurements were performed in a previously described

3. Results and discussion

The kinetic energy of the condensing particles seems to play the crucial role for film growth of rare-earth

0-

0

.

00

§"

a]"

~ 0 ~

0

-'

~

(a)

§" -'

I

00

-:::.

.

a,'''

1317

0-

.

~

0

&

rtf

w.J

N

'"

.

~

&

~

m~

w~

w..

-'

-'

-'

.

-'

..

I

0 0

0

:::.

m~

(b)

0

u"

~

0 0

§"

...... en

~

rtf u"

:=: C

-

...>... ::J

0

00

::;

0

~

&

m~

u"

ttl

:c.....

0

;:;;

m"

u"

rtf u

0

N

rtf

m" E

'"~

~

.

~

+-'

0

ttl ........

>-

'wC

(c)

+-'

g

0

0

rtf

rtf

(/)

(/)

E

Q)

+-'

0

~

~ 0

-

E

C

00

::;

'"~

& m" E

~

E

(/)

C/)

(/)

0

0 0

~

0

::;

0

~

(d) 0

g

§"

w..

-af

m~

>

>

.A>

I-I

I 20

30

40

50

60

0 0

~ m~

>

70

0-

0

&

~

af

m~

>

80

>

90

diffraction angle 28

~

'"

'"~

rtf >11.

>

100

w..

110

120

[0]

Fig. 2. X-ray diffraction spectra of sputtered rare-earth hexaboride coatings. Static deposition; substrate, molybdenum; chromium KIX radiation; sputtering power density, 4.8 W cm -2; argon pressure, 3 Pa; bias voltage, -100 V; substrate temperature, 200 dc. (a) LaB 6 -based coating; (b) CeB 6 -based coating; (c) 5mB6 -based coating; and (d) YB 6 -based coatings.

IV Waldhauser et al. " Surface and Coatinl(s Technolol(Y 98 (1998) 1315 -1323

1318

atomic radius: • Y • La & Ce 'Y Sm

2.30

sum of XRD peak heights'

enthalpy of formation'

o

YB s

l::.. CeBs I

0 'V

LaBs

ED

5mBs

~ CeBs

YBs

900

:0 ...as

2.00

0

iii

'"

l::..

~

(J

'E

-200

0 1.90

1.80

170 I-

•

-300

&

•

-600 -700

ED

40

I

45

50

55

0

E

....,

~ co

Ol N

0

...:

J:
'Y

I

-

-400 -500

65

60

~[§J

-800

Vi

0-

(J ~

'V

til :J

800

-100

radius range of stable ReB 6 compounds

2.10

5mB s

0

~

2.20

~

LaBs

100

I

EEl

0

EEl '"

700

~

(/)

.c 600 C) 'Qi .c 500

.::.t.

400

0-

300 200 100

as

Q)

0

cr: X 0

E :J

(f)

a

[§J

atomic number

Fig. 3. Atomic radius range of stable hexaborides. radii for different rare-earth elements. relative crystallinity of corresponding sputtered coatings and thermodynamic stability of the hexaboride phase expressed by the standard molar enthalpy of formation.

hexaboride PVD coatings [10,17-25]. Thus, the argon pressure affecting the scattering of sputtered energetic target atoms [26] showed, among the deposition parame• ters the strongest influence on the film structure and properties. Films deposited in the static mode at argon pressures ranging from 2.5 to 3.5 Pa, at substrate bias voltages between - 50 and - 100 V and at a substrate temperature of 200 °C were dense, smooth and adherent. Lower or higher pressures resulted in flaky peeling of the coatings corresponding to the reports of Mroczkowski [10]. Static deposition at a sputtering power density of 4.8 W em - 2 resulted, for all hexabor• ides investigated, in growth rates of 40-50 nm min -1. The typical morphology of coatings deposited by magne• tron sputtering from the targets used is shown in the SEM micrographs in Fig. I. The variation of the depos• ition parameters argon pressure. bias voltage and sub• strate temperature indicated that in the case of static deposition a dense extremely fine-columnar structure [see Fig. I (b)] with a smooth coating surface is formed at conditions with moderate energetic contribution for film growth. In the complementary growth region, i.e. extremely low or high energetic contribution (e.g. high argon pressure, low substrate temperature or high bias voltage), the structure of the coatings becomes extremely fine-grained to fracture-amorphous [see Fig. I (a-d)]. La-B, Ce-B and Sm-B coatings deposited onto rotating substrates exhibit generally a fine-columnar structure showing a micro-rough and micro-porous surface

described and discussed earlier [17,20]. The rotation of the substrates results in a discontinuous flux of sputtered target atoms and argon ions depending on the alignment of the substrate surface to the target. This may decrease the average energetic distribution for film growth favour• ing columnar growth [27]. Static as well as dynamic deposition of Y-8 coatings yielded a fracture-amor• phous film structure. All Re-B films deposited were overstochiometric. Increasing the argon pressure caused a gradual shift in the chemical composition of the coating from B:Y = 7.5, B:Ce=7.2, B:Sm=6.8 and B:La=6.7 at I Pa to B:Y = 6.9, B:Ce=6.8, B:Sm=6.4 and B:La=6.3 at 4 Pa. For non-reactively sputtered films, stochiometric deviations of the coating from the target composition are related to interactions of the sputtered target atoms with the plasma discharge in the transport phase and at the surface of the growing film, assuming that diffusion processes in the target are inhibted [28]. Therefore, the main reasons for the observed influence of the argon pressure on the chemical composition of the coatings are collisions of sputtered energetic particles with argon atoms. These collisions may be characterized by both the collision cross-sections and energy transfer coeffi• cients. Special notice should therefore be paid to the particle radii determining the collisions cross-sections and to the atomic masses determining the energy transfer coefficients [29]. The general overstochiometry of the coatings deposited may therefore be interpreted by the

W Waldhauser et al. / Surface and Coatings Technology 98 (1998) 1315-1323

high mobility of the small boron atoms within the discharge thus arriving comparatively easily at the sub• strate, especially at low argon pressures. With increasing argon pressure, the scattering of boron atoms by argon is increased, resulting in efficient thermalization and reduced deposition probability. In contrast, rare-earth atoms are not so strongly affected by scattering owing to the increasing argon pressure because of their high atomic masses with respect to argon. This has been confirmed by Monte-Carlo simulations of the behaviour of sputtered lanthanum and boron atoms [30], which show that with increasing working gas pressure an increasing number of boron atoms is thermalized. Typical XRD traces of coatings deposited in the static mode onto molybdenum substrates are shown in Fig. 2. The relative peak intensities given for randomly orien• tated ReB 6 and molybdenum (corresponding to the JCPDS powder diffraction file [31]) have been corrected for the chromium KO( radiation [32]. The texture coeffi• cients TC(hkll [33,34] J

to a high thermodynamic stability of the hexaboride phase expressed by the standard molar enthalpy of formation ~H~. 298' Calculations of the lattice parameters using the (100) and (110) peaks [Fig.4(b)] indicate the existence of compressive stresses in the films [10, 19,23]. Static depos• ition onto tungsten substrates at an argon pressure of 3 Pa resulted in lattice parameters obtained from the (100) peaks of a LaB6 =4.262, a CeB6 =4.180 and aSmB6 = 4.207 A, corresponding to distortions of +2.62, +0.94 and + 1.86% with respect to the standard values of aLaB6 =4.153, aCeB 6 =4.141 and aSm B6 =4.130 A [31]. Coatings deposited at corresponding sputtering parame• ters onto rotating substrates showed lattice parameters of aLaB 6 =4.212, aCeB 6 =4.178 and aSm B =4.191 A. result• ing in lattice distortions of + 1.42, + 0.89 and + 1.48%. The differences in the lattice parameter between static and dynamic deposition are interpreted by the possibility of stress relaxation at voided grain-boundaries [17] resulting in decreasing distortion. Corresponding to earlier work on LaB 6 -based coat• ings [21, 23], the distinct (l 00) texture of the LaB 6 and 6

(1) 3.0

0' o .....

where I and 10 are the intensity values for a peak (hkl) of interest taken from the XRD traces and the powder diffraction file, respectively, and n is the number of peaks considered, were calculated for coatings deposited onto molybdenum and tungsten substrates. It should be noted that a random orientation would lead to TC values of unity, whereas a complete dominance of one peak would return the number n of peaks considered. In all La-B and Ce-B films investigated, the formation of the cubic ReB 6 phase exhibiting a distinct (100) texture [Fig. 2(a and b) and Fig. 4(a)] could be detected. Sputtering from the 5mB 6 target led to a lower crystall• inity of the coatings and to a nearly random orientation of the 5mB 6 phase [Fig. 2(c) and Fig.4(a)]. The Y-B films showed nearly amorphous film growth [Fig. 2(d)). The three peaks of low intensity not marked in Fig. 2(d) represent the KfJ peaks of the molybdenum substrate. Considering geometrical aspects, i.e. the size of the rare• earth atom has to be appropriate to the metallic sites in the rigid boron skeleton [1-3], the metallic radii for the rare-earth atom may vary in the range between 1.75 and 2.23 A (Fig. 3) in order to form the corresponding hexaboride phase [7]. Following the different radii for lanthanum (1.870 A.), cerium (1.825 A.), samarium (1.802 A.) and yttrium (1.776 A.), a decreasing tendency to stabilize the crystalline hexaboride phase in sputtered films is observed which may be described by the sum of the (100), (110) and (111) peak heights of the corre• sponding XRD traces (Fig. 3). Additionally, Fig. 3 shows that a high crystallinity of the coatings correlates

1319

U I-

2.5

r-.---.---,-..--..........-r--r---.-........,,....-..--....--r---. -o-LaB6 - _ 6- CeB

6

_.v-. 5mB6

2.0

,

: t

:

: : +___ _ ;

,

: ,

,

: ,

1.5

--

--"---------+---

0.0

--

--f (a)· j--.------ --~---- ------- :----- ------r----------+-------

4,

::;;-~-~-~-o-o-~-~.~-~-~-~-~---

:.: • • ·--·-i~=::t-~;:E~_~t:::_E·~-.I. ,

,

1.0

~ 0'

"

15

2.0

2.5

3.0

,

-~---

35

4.0

4.30

r-.----.---,,....-..--....--r-y---r---r........,,....-.,.---.---,---.

4.25

----i5:.:..:.:-----+---------j----------+----------~--------- A- CeB6

:

,

i

:

:

-D-laB6

:

'-0

'

,

-'''V--SmB6

lii:':~-~':'-: bulk CeB; 0. ! bulk Sm~6

:

.....

_

i :::t--I~J=J-~-~~ ~

~

4.1 0

""----r----------j-----------j----------+--------+----------:----------+--

4.05

l-...L-...............J_...........I-........_ l - -.............L.--'

: (b)' 10

1.5

: 2.0

2.5

3.0

...............--L---I

3.5

40

argon pressure [Pal Fig. 4. Texture coefficients (a) and lattice parameters (b) determined from the ( 100) and ( 110) peaks. respectively. of the hexaboride phase in coatings deposited onto molybdenum substrates as a function of the argon pressure. Static deposition: sputtering power density. 4.8 W cm - 2; bias voltage. - 100 V; substrate temperature. 200°C.

W Waldhauser et al. / Surface and Coatings Technology 98 ( 1998) 1315-1323

1320

CeB 6 phase becomes even more dominant with increas• ing argon pressure [Fig. 4(a»), whereas the lattice param• eters calculated from the diffraction peaks decrease with increasing argon pressure [Fig. 4(b»). This behaviour can be interpreted by the increased number of collisions of the sputtered energetic particles with argon atoms resulting in a larger number of thermalized atoms [35] and a decreased number of argon atoms incorporated interstitially in the ReB 6 lattice. The pressure-indepen• dent texture coefficients and lattice parameters of 5mB 6 films (Fig. 4) can be explained by the compara• tively high atomic mass of the sputtered samarium atoms being not as efficiently thermalized by collisions with argon atoms at higher pressures. Fig. 5 shows the dependence of the lattice parameter of coatings deposited in the dynamic mode onto polished tungsten substrates on the annealing temperature after deposition. Owing to the possibility of stress relaxation, the lattice distortion decreases with increasing annealing temperature. XRD traces of coatings sputtered from the YB 6 target showed, after heat treatment at 900 DC, peaks of low intensity indicating the formation of the hexaboride phase and allowing the determination of the lattice parameter given in Fig. 5. All coatings deposited onto rotating tungsten sheets were adherent throughout annealing at temperatures of up to 700°C. Obviously, owing to thermal stresses, Ce-B and Sm-B deposits tended to slightly flaky peeling after annealing for 30 min at 900°C during heat treatment or cooling down, respec• tively, whereas Y-B films showed a very poor adherence probably caused by stresses due to the formation of the

YB 6 phase. On the contrary, coatings deposited dynami• cally onto tungsten filaments were adherent throughout emission testing at temperatures of up to 1400 dc. This is attributed to the relative high ratio between coating and substrate thickness allowing the relaxation of ther• mal stresses. Before emission testing, the heating current was increased slowly and the cathode surface was activated by maintaining an annealing temperature of the coated filaments of 900°C for 30 min. It can be seen from Fig. 6(a) that coated filaments work at temperatures approximately 1000 °C below the operating temperature of uncoated tungsten filaments while emitting nearly the same electron current. The values of the thermionic work function at temperatures ranging from 900 to 1100 °C were calculated from Richardson plots as tPLaB6 =2.64eV, tPCeB 6 =3.27eV, tPsmB 6 =3.l1eV and tPYB =3.05 eV. The linear correlation in Fig. 6(b) indi• 6 cates that among the crystalline hexaboride coatings investigated, low work function values correspond to high lattice parameters. Similar to the crystallographic dependence of the work function of LaB 6 single crystals [36], this behaviour can be interpreted by varying rela• tive concentrations of rare-earth atoms in the outermost surface layer as well as different dipol contributions and surface reconstructions determined by the lattice parameter. Lifetime tests were performed in the described diode test cell [17] at a residual pressure of 5 x 10- 4 Pa and at a constant average cathode temperature of 1100 DC. Fig. 7 shows the emission current of coated tungsten

annealing temperature [DC] o 4.24 - 4.22

0

0

0

0

0

: -

-

-

-

o

~

-

-

-

-

-

__ -

-

~ __ -

400

600

I : -

-

-

0

0

_

0

0

~

0

:

:

:

i

~----------------

O-!

...

-E

200

I

:

: 0

0

0

0

_

800

1000

I :

:

~

~--

:

:

:

: - 6.- CeB6 (100) +__

:

i i i

-0-LaB 6 (100)

--v--

8mB 6 (100)

Q)

... Cl3 Cl3

a.

~

~

4.18 4.16 4.14

_

bulk LaB

6

i

:

:-

-:- ___

:

~-------.. bulk CeBe --------..----)------------------)---....-------- ..-...~ -->.<-;-~)--------~

4. 12

bulk 8:mB e ------------------ -

4.10

--

200

0 -- ---.0

i i i

'v i

-i--------------..-----i---------------------i----------.----------i------------.--------i-o. -----.

0,

bulk YBe ---------- --------l------------- --------l--------..---- -------l------------------ --l--

:

400

i i i

600

800

1000

0

_

1200

annealing temperature [K] Fig. 5. Lattice parameters of the hexaboride phase in coatings deposited onto tungsten substrates as a function of the annealing temperature after deposition. Dynamic deposition; sputtering power density, 4.8 W cm -2; total pressure, 3.0 Pa; bias voltage, - 50 V; substrate temperature, 200°C.

W. WaJdhauser et aJ. I Surface and Coatings TeclmoJogy 98 (/998) /3/5-/323

1321

60 D

(a)

.......

« ~

-...... C

Q)

50

....-

40

-----_ -- .. _--------_

30

:-..--. _._._.. . . "-'--r" - ,

,

--_. __ ._--_.....-- ----_ ..... --------

- [ J - filament

coated with La-B - 1::>.- filament coated with Ce-B --'\1-' filament coated with 8m-B •.<>-•. filament coated with Y-B -0- uncoated tungsten filament

:::::::::,::::;;1>>:::r:::::::::[::I::

;:]

0

c 20 .2 en

. . . . . . . . . . . ,:tL/>.f'!t ..j

.~

E 10 Q)

~~o-~~".,

0

~oo<>,,~t 1600

1200

1000

800

....

1800 2000

2200

temperature [0C] 3.4

D

.......

>Q)

--

I::>.

'il

La-B coating Ce-B coating 8m-B coating

3.2

c: 0

-... +: 0

c

3.0

...

;:]

~

0

:::

2.8

0

'c 0

2.6

-

2.4

'E...

D

Q)

.c:

4.17

(b) 4.18

4.19

4.20

4.21

4.22

lattice parameter of hexaboride phase [A] Fig. 6. Electron emission characteristics of coated and uncoated tungsten cathodes: (a) emission current as a function of the operating temperature; (b) dependence of the work function on the lattice parameter.

hairpin cathodes (coating thickness 211m) as a function of the operating time. Owing to activation processes [10] determined by the evaporation of pollutions such as oxides [37] and a shift in the chemical surface concentration [6], the emission current increases at the beginning. As a result of its low work function, the La-B-coated cathode shows the highest electron emis• sion. However, after 80-110 min, the first damage occurs at the tip of the cathodes and the emission current starts to decrease. The geometry of hairpin cathodes causes an inhomogeneous temperature distribution with extremely high temperatures at the tip of the cathode

resulting in local melting and evaporation of the coating material (Fig. 8). In agreement with the results obtained in the diode test cell, the electron emission during testing of an La-B-coated hairpin cathode in the SEM starts to decrease after an operating time of about 90 min. Cathodes with straight filaments used for the determina• tion of the work function values [17] showed a uniform temperature. At an operating temperature of 1400°C, the emission current of these filaments was stable for approximately 3-4 h. Using a lifetime diagram of LaB 6 single crystal cathodes [38], an evaporation rate of 0.58 I!m h -1 was determined for a temperature of

II

«

2;

.....C

II itltlli,'IIS('1'

,'I

til.

,\lIl'ltI('(' tllltl ('otll jllgs 'f(', 'lillologr

'is ( I 'NS I 1315 13:!3

10

Q)

'•

'-

::::l

U C

o

CIl CIl

E

Ql

O. 1 L_-L-----J'-----L--l.~LU...L__

1

___l.._.l.._.L...JL_LJ...Lu....._

__l_

100

10

_L..._L...L...L....L.l..LL.__.J

1000

operating time [min] h~. 7. Fmi"j"n currcnt "I' «)ated tun~sten hairpin cathodes as a function of the operatjn~ time. Average cathode temperature. 1100 C: residual pre"ure. ;; " 10 4 Pa.

the above-mentioned diagram and can be attributed to the low cathode temperature of 927 C and the low pressure ( I x 10 ' Pal used.

4. Conclusions

hl!. X. SI-M micrographs "I' the tip or a La B-«)atcd tun~sten hairpin cathode after practical use in ~I SEM. Testing time. ~ h: accclerating \oltage. ~o k V: heating current. ~.07 A; ahsorhed electron currelll. ~()O pA.

1400 C and a residual pressure of 5 x 10' 4 Pa. Assuming a stable emission of the coating until its complete evapora• tion. the observed live time is in good agreement with the evaporation rates of single crystals. Thus. the lifetime of coated cathodes at temperatures of up to 1400C seems to be limited by evaporation processes [17] deter• mined by the quality of the vacuum. At higher temper• atures. the stability of the boride films is influenced by chemical reactions with the substrate [39]. The very long lifetime of \000 h of an LaBh-coated filament (film thick• ness I ~m) reported by Mroczkowski [10] agrees with

Single-phase ReB" (where Re = La. Ce. Sm) overstoch• iometric films with fine-columnar to extremely fine• grained structure were prepared by non-reactive DC magnetron sputtering of ReB(, targets in static and dynamic mode at argon pressures ranging preferable from 2.5 to 3.5 Pa. Y B films showed nearly amorphous film growth. Among the boride coatings investigated. LaB h and CeB h arc found to be the most promising emissive material for coated cathodes. Owing to thermal stability and evaporation behaviour. the application of ReBh-coated cathodes seems to be limited to low opera• tion temperatures and residual pressures.

Acknowledgement This work was supported by the Bundesministerium fUr Wissenschaft. Verkehr und Kunst (Austria).

References [1] J. Etourneau. P. Hagcnmuller. Phil. Mag. S~ (.i) ( 19X5) SXlJ. J. Etourneau. J. Uss-Common Met. 110 (llJRS) Iti7.

[~]

W. Waldhauser el al. I Surface and Coalings Technology 98 (1998) 1315-1323

[3) K.E. Spear. in: A.M. Alper (Ed.). Phase Diagrams: Materials Science and Technology. Academic Press, New York. 1976. p. 91. [4) J.M. Lafferty, J. App\. Phys. 22 (1951) 299. [5) L.W. Swanson. D. McNeely, Surf. Sci. 83 (1979) 11. [6) M. Futamoto. M. Nakazawa, U. Kawabe. Vacuum 33 (10) II) 12) (1983)727. [7) F. Binder. Radex-Rundschau I (1977) 52. [8) L.J. Favreau, Rev. Sci. Instrum. 36 (1965) 856. [9) R.S. Khairnar. P.W. Mahajan, D.S. Joag, A.S. Nigavekar. P.L. Kanitkar, J. Vac. Sci. Techno\. A 3 (2) (1985) 398. [10) S.J. Mroczkowski. J. Vac. Sci. Techno\. A 9 (1991) 586. [II) A. Yutani. A. Kobayashi. A. Kinbara. App!. Surf. Sci. 7071 (1993) 737. [12) Y. Okamoto, T. Aida, S. Shinada, Jap. J. App\. Phys. 26 ( 1987) 1722. [13) T. Nakano. S. Baba. A. Kobayashi, T. Kajiwara, K. Watanabe. J. Vac. Sci. Techno!. A 9 (1991) 547. [14) J.G. Ryan. S. Roberts. Thin Solid Films 135 (1986) 9. [15) G.A. Meerson. N.N. Zhuravlev. R.M. Manelis, A.D. Runov. A.A. Stepanova. L.P. Grishina, N.V. Gramm. Inorg. Mater. 2 (1966) 527. [16) S.V. Ermakov. B.M. Tsarev, Radio Engng Electron. Phys. 10 (1965)833. [17) W. Waldhauser, C. Mitterer. 1. Laimer. H. Stari. Surf. Coat. Tech• no\. 7475 (1995) 890. [18) M. Mushiaki. K. Akaishi, T. Mori, Y. Kubota. Y. Funato, O. Motojima. Mater Sci. Engng A 163 (1993) 177. [19) T. Kajiwara. T. Urakabe. K. Sano. K. Fukuyama, K. Watanabe. S. Baba, T. Nakano, A. Kinbara. Vacuum 41 (4 )5)6)( 1990) 1224. [20) W. Waldhauser. C. Mitterer. A. Winkler. Le Vide: science, tech• nique et applications 279 (1996) 128. [21) J. Stallmaier. C. Mitterer, J. Barounig. Le Vide. les Couches Minces 261 (1992) 265. [22) W. Waldhauser, C. Mitterer, P. Schmalz. H. Stiiri. J. Barounig. in: H. Bildstein and R. Eck (Eds.), Proceedings of the 13th

1323

International Plansee Seminar, Metallwerk Plansee. Reutte. 1993. vo!. 4. p. 267. [23) C. Mitterer, J. Komenda-Stallmaier. P. Losbichler. P. Schmolz. H. Stari, Surf. Coat. Techno!. 7475 (1995) 1020. [24) C. Mitterer, J. Komenda-Stallmaier. P. Losbichler. P. Schmalz. W.S.M. Werner. H. Stori. Vacuum II (1995) 1281. [25) C. Mitterer, J. Komenda-Stallmaier. P. Losbichler. P. Schmalz, W.S.M. Werner. H. Stori. in: R.W. Lang, M.A. Erath (Eds.). High Tech in Salzburg - Creativity in Advanced Materials and Process Engineereing, SAMPE, Niederglatt. 1995. pp. 141-152. [26) I. Petrov, !. Ivanov. V. Orlinov, J.-E. Sundgren. J. Vac. Sci. Tech• no\. A 5 (1993) 2733. [27) J.A. Thornton. Ann. Rev. Mater. Sci. 7 (1977) 239. [28) G. Betz, M. Opitz. P. Braun. Nuc\. Instrum. Meth. 182183 (1981) 63. [29) J.A. Thornton, in: R.F. Bunshah (Ed.). Deposition Technologies for Films and Coatings. Noyes. Park Ridge. NJ. 1982. pp.170-243. [30) A. Kinbara, T. Nakano. A. Kobayashi, S. Baba, T. Kajiwara. App\. Surf. Sci. 7071 (1993) 742. [31) Powder Diffraction File. Joint Committee on Powder Diffraction Standards. International Center for Diffraction Data. Swarthmore. PA. Card 6-0401, 11-0670,36-1326. 16-0732.4-0809. [32) B.D. Cullity. Elements of X-ray Diffraction. Addison-Wesley, Reading, MA, 1978, p. 131. [33) A.J. Perry, J. Schoenes. Vacuum 36 (I )2)3) (1986) 149. [34) G.B. Harris. Phi\. Mag. 43 (1952) 113. [35) R. Elsing. Surf. Coat. Techno!. 49 (1991) 132. [36) M. Gesley. L.W. Swanson. Surf. Sci. 146 (1984) 583. [37) R. Nishitani. M. Aono, T. Tanaka. C. Oshima, S. Kawai. II. Iwasaki. S. Nakamura. Surf. Sci. 93 (1980) 535. [38) Balzers, Elektronenmikroskopische Praparationstechnik, Balzers, 1986. p. 169. [39) C. Oshima. S. Horiuchi. S. Kawai. Jap. J. App\. Phys. 2 (I) (1974) 281.