A N N A L E S
U N I V E R S I T A T I S M A R I A E C U R I E - S K L O D O W S K A L U B L I N — P O L O N I A
VO L. X L V I/X L V II, 9 SE C T IO AAA 1 9 9 1 /1 9 9 2
In st itu te o f Physics. M. Curie-Sklcxlowska University
L. G Ł A D Y S Z E W S K I
M ass S p e ctr o m etr ic In vestig a tio n s o f th e A lkali M igration on T u n gsten at Low C overages
The surface diffusion for five alkali metals on polycrystalline tungsten and ionic thermal desorption are studied by a method based on the alkali ion current noise arising from the fluctuation of the work function as a result of random fluctuations of the alkali adsorbate density.
The activation energy for surface diffusion for Li, K, Rb and Cs have been determined by measuring the spectral density functions and their parameters.
M E T H O D
The method is based on the relationship between the number n of adatoms adsorbed on a small region of the ions emitter and the ion thermoemission current i emitted from this region.
Due to random surface migration of adatoms, n undergoes fluctuations dn around its mean value. These fluctuations are reflected in current fluctuations (//, with a spectral density S: S = (di~)/df, [1].
The mean square of the ion current fluctuations is given by:
(di2) = (5(llfe /,/t„kTl)2n0DT0, (1)
where /0 is the DC ion current intensity, /* — the dipole moment of the adsorbed atoms, / — the diffusion length, D — the diffusion coefficient, iiq — the concentra
tion of adsorbed atoms and tq is the mean residence time of adatoms on tungsten [2.3,4,5],
To investigate the diffusion, the so-called ” normalized mean square” of the ion current fluctuations was applied:
T ab le 1. T h e su rface diffusion en ergies for alkali e le m e n ts (E: a c tiv a tio n en erg y for su rface d iffu sion [eV], r t : ion ic radius o f the a d a to m [A])
A d atom r. [A] E [eV] on W Li 0 .6 0 0 .6 7 ± 0 .1 0
N a 0.95 0 .5 7 ± 0 .0 5
K 1.33 0 .4 1 ± 0 .0 2
Rb 1.48 0 .4 4 ± 0 .0 5
Cs 1.69 0 . 2 9 i 0 . 0 5
where E is the activation energy for surface diffusion [2].
Value tq was estimated using the autocorrelation technique [2,3,9].
E X P E R IM E N T A L
The measurements were performed in a stainless steel vacuum chamber allowing a high vacuum of 10“ 9 Torr. The procedure of depositing atoms on the emitter was described earlier [3,4] by the author, and details concerning the ion source construction can be also found there.
The anode of the thermal emission of ion source was a tungsten ribbon of size 10 x 0.8 x 0.025 mm.
The ion current fluctuations were amplified by a wide-band electrometer.
The spectral density functions were investigated with a Unipan-237 selective nanovoltmeter with relative selectivity of A / / / = 0.014.
R E SU L T S A N D D ISC U S SIO N
The spectral density function can be approximated by the Lorentzian function:
S = So/[1 + (u/to)2], (3)
where To is the mean residence time of the adatoms on tungsten surface and
U) — 27t/ .
It is well-known that the relationship between the {dir) and the spectral density function is given by:
r e v
( d f ) = S ( f ) d f , Jo
and rCO
(di7) = S0/[l = (u r0)2](// = 50/4 r 0.
Jo
The low frequency spectrum is flat for /(l/2 7 rr0. For / — 0, S So and its asymptotic value is:
50 = (2?0e///^o^T/)2n0D r02.
The turnover frequency (3dB) fo — 1/2tttq and the high frequency asymptote is proportional to / “ 2.
F ig. 1. S p ectra l d en sity function S ( J ) for p o ta ssiu m on tu n g ste n for tem p er a tu res 1270 K and 1360 K
5040
T [ K - 1]
F ig. 2. N orm alized m ean square fluctu ation as a fun ction of the reciprocal te m p er a tu re
Edir. [eV]
r, [ A ]
F ig . 3. Surface diffusion energy o f alkali a to m s on p o ly cry sta llin e tu n g ste n as a fu n c tio n o f ion ic radius r t . (M o - M o r i n [7], G - G o i n e r [10], Be - B ę b e n [l 1 ], P - P o p p [12], K - K 1 e i n t
[13], D - D ą b r o w s k i [ l l ] , B - B a y a t [l 5], B1 - B 1 a s z c z y s z y n [16], .1 - J a m b a [17])
411,
39 L
39.,
LA. J
N T M
F ig. la. M ass sp e c tr a o f th e p o ta ssiu m (D C co m p o n e n ts): "N" — n atu ral sa m p le , "T" — tracer,
” M ” —■ m ixt ure
The surface diffusion energies were determined from the thermal dependences of normalized mean square fluctuations in (see Fig. 2).
It was determined that the values E versus ?*, (ionic radius) represent a straight line (Fig. 3):
E — 0.88 — 0.33?%.
In Fig. 3 the data obtained in these experiments are marked with full circles.
For the sake of comparison some values of E, obtained by other authors are also included.
J
F ig. -lb. M ass s p e c tr a of th e p o ta ssiu m ob ta in ed for n o ise c o m p o n e n ts (sa m p le ” M ” )
The crystallographic structure of the tungsten ribbon was examined using the X-ray diffraction technique. From those one can see that (001) planes dominate, it means the polycrystalline ribbon reveals a (001) texture.
From the diffraction spectrum we can estimate the percentage contribution of various planes. The spectrum was compared with data for a powder sample.
An additional isotope fractionation effect was also reflected in the noise mea
surements. The isotopic ratio for potassium and lithium could be determined in these experiments by the analysis of noise components corresponding to the iso
topes [3]. Measurements were performed on a magnetic mass spectrometer with 90°
deflection and resolution of 300.
A mixture of two potassium samples: one of them was a normal isotopic composition (39K: 92.96%, 41I\: 7.03%), the other was a potassium tracer (39Iv:
4.2%, 41K: 95.8%) was used in isotopic fractionation measurement. The mixture was prepared by isotope dilution technique [5,6].
F ig. 5. T em p eratu re d ep en d en ce of the norm alized m ean square m e x p e r im e n ta lly o b ta in e d for 30K (solid lin e) and ca lcu la ted for 41 K (d ash ed lin e)
The following isotopic composition was obtained for mixture: (39K: 47.64%, 41K:
52.36%).
For two isotopes of potassium in sample M, the isotopic ratio ?o obtained by measurement of DC ion current component: 7*0 = = 1.072 whereas the isotopic ratio r n, obtained by measurement of noise component (cf. eg. (1)):
»•., - ((<K?,)/(</&))1/2 = r o y j D ^ / D ^ r ^ / T i , ,
rn = 1.05. ________
Since = a [7,8] and a = (A/3c>, A/41 are a mass of the isotopes: for potassium: a = y/Ą 1 /39 — 1.025, thus:
Dm/ Dąi = (ro/rn)2 • o.
Let a = j'o/ r ny this value will be called the ’’additional fractionation factor” , therefore:
D:vj/D.\\ — a '(ij.
For investigated potassium sample a\ — 1.02 then D3{t/D4[ = 1.07 for T =1640 K.
These results signified that for two potassium isotopes we observed two values of the surface diffusion energy:
D39/ D 4i = [^39exp( — £39/kT)]/[a4[ e x p ( - E4i/kT)] = a • exp[(£4i - E30)/kT].
Our measurement give: £ 3 9 = 0.41 eV, therefore £41 = 0.43 eV.
Similar measurements and calculations performed for natural Li samples give:
7*0 = ij/i$ = 11.42 ±0.08, rn = (dil))1/ 2 = 9.5 ± 0.2 and D^/D-j — 1.5 (for T =2040 K), therefore: £ 7 = 0.67 ± 0.10 eV [1] and £g = 0.61 ± 0.10 eV.
This work was supported by the 2298/2/91 grant.
REFERENCES
[1] G ł a d y s z e w s k i L., Surface Set., 213 (1 9 8 9 ), 481.
[2] G ł a d y s z e w s k i L., Surface Sci., 231 (1 9 9 0 ), 120.
[3] G ł a d y s z e w s k i L., Is ot openpraxi s , 22 (1 9 8 6 ), 212.
[4] G ł a d y s z e w s k i L., Surface Set., 200 (1 98 8), 386.
[5] C o h e n K. , The or y of Is ot ope Se p a r a t i o n, M e G raw -H ill B ook C o m p ., N ew York 1951.
[6] W e b s t e r R. K ., [in:] Ad v a nc e s in Ma s s S p e c t r o me t r y , J. D . W aldron (e d .), L on d on 1959.
103.
[7] G o m e r R ., Sol i d St at e P h y s ., 30 (1 9 7 5 ), 93.
[8] M o r i n R. , Surface Sci., 155 (1 9 8 5 ), 187.
[9] M ę c i e w s k i R. , Ac t a Vniv. Wrat i sl av. , 10, 147 (1 9 7 1 ).
[10] S c h i n i d t L,, G o m e r R ., J. Che m. Phys. , 45 (1 9 6 6 ), 1605.
[11] B ę b e n J., K 1 e i 111 C h ., M ę c l e w s k i R. , J. P h y s ., C 9, 45 (1 9 8 4), 13.
[12] P o p p G. , Ann. Phys. , B 13 (1 9 6 4 ), 115.
[13] K l e i n t C h ., Surface Sci., 25 (1 9 7 1 ), 394, 411.
[14] D ą b r o w s k i A. , K l e i n t C h., Surface Sci., 119 (1 9 8 2 ), 118.
[15] B a y a t B ., W a s s m u t h H. K. , Surface Sci., 133 (1 9 8 3 ), 1.
[lG] B l a s z c z y s z y n R. , K l e i n t C h., J. Phys. , C 2, 3, 47 (1 9 8 6 ), 393.
[17] J a m b a D. M. , H u s m a n n O. K. , J. Appl. Phys. , 38 (1 9 7 6 ), 2630.
