Intermetallic compound formed by electrodeposition of indium on antimony

Intermetallic compound formed by electrodeposition of indium on antimony

Journal cd Alloys EISEVIER 0~~5-X3XW/Y7/$17.(po PII SOY2S-#.3X8( (51 lYY7 Y7 )0009(1-0 Elhevter Scicncc S../\ 411 rtg’lt~ axi Compounds 25Y ...

2MB Sizes 8 Downloads 20 Views

Journal cd Alloys

EISEVIER

0~~5-X3XW/Y7/$17.(po PII

SOY2S-#.3X8(

(51 lYY7 Y7 )0009(1-0

Elhevter

Scicncc

S../\

411 rtg’lt~

axi

Compounds 25Y

rwrved.

( 1YY7 1 23-l-240

A different method of preparing InSb thin films was proposed by Mengoli et al. ( 111 involving a sequential deposition from chloride baths, Sb first, then In. In was deposited at -0.268 V vs. In from a bath containing 0.3 PvI InCl, and 1 M KC1 whose pH was adjusted to 1.5-2.0 with HCI. The two-layer samples, consisting of an inner Sb and an outer In layer (about 1 pm each) deposited on a Pt substrate, were then annealed at 150, 175 and 185 “C for times from 1 to 15 h. Their percent conversion to InSb was determined as a function of time by stripping voltammetry in a strongly acidic solution ( 1.7 M HCI, 1.6 M H,SO,) and the results were qualitatively confirmed by X-ray diffraction (XRD). Only the In and Sb peaks were detected after annealing at 150 OC. a temperature close to the In melting point ( 150.6 “C) while buih the typical peaks of InSb and of the unreacted components were present at 175 OC. their relative amount depending on the annealing time. The single components peaks completely disappeared by annealing at temperatures equal to or higher than 185°C for 5-10 h. Hobson Jr. et al. [ 121 showed that t e rate of f~~rnli~tiol~ of InSb at the interface between In and Sb at 1 “C is elecgreatly increased when a composite electrode of Sb is made the cathode in an electrolytic diffusion coefficient was lower than 2x uring the thermal synthesis of InSb and l.SX 10 ‘(’ mJ s ’ e dt the time of ncxi that the data avai aper were insufficient t diffcrencc in the diffllsi~~~~ rate. f-diffusion in lnSb WilS invchtigale L\ traxr lechnique illld tk iKtiVil

vacancy mecl~anistn in this c mechanism for seif-diffusion accounting for above data seems to be vacancy diffusion in each df the face-ccntred cubic sublattices formed the two ~~9l~stitil~~~ts4.91 th ~~)nsid~~~ti~9~~s. th diffusing

entities in

electrode on the bottom. Al! the performed in galvanostatic conditions tiostat-galvanostat) with a 668/RM Amel potentiostat and re (868 Amel). The potential of the Sb elec

were such that no by monitoring the el circuit. Another set of

Thick ( 18 pm) deposits wecI: rough and imperfect with typical holes but the Pn dot map analysis of their crosssection showed their prac:ical continuity. SEM-EDS was performed at several points of this section from the surface towards the bulk of the specimen. An example is shown in Fig. l(a) for a deposit observed two years after electrodeposition. The analysed points are labeiied with progressive numbers. They show (Fig. l(b)) the continuous decrease in the In atomic percentage which reaches a value close to 50% at the deposit-substrate interpdce. thus indicating the formation of InSb. At higher temperatures, the deposits were continuous (Fig. I(c) shows a topographic view) and it was easier to observe InSb. for example after t~~~~t~n~the deposit for 334 h at I 10 “C (Fig. I(d) shows the sample cross-section). This sample was scanned as to In ~~~nlposition atong the white line in the micrograph

and the In-line protile was reported below the SEM image. Although such a profile was difficult to read because of the skewness of the surface. the comparison of the ratio between the height of the peaks with that between the theoretical In/Sb concentration showed that the sample had the three-layer structure InlInShlSb from the surface to the bulk. The XRD pattern of the thicker deposits practically showed only the In peaks. Indeed. X-rays were completely absorbed in the external part of the deposit and could not reach the In/InSb interface, the intermetallic compound lying beyond their penetration depth. Fig. 2 compares the schematic AST and InSb 1141 with the XRD pattern f a thin (0.8 Frn) deposit heated at I35 “C fo hug h smic Iines are superimposed in the AS it was possible to

where !,, and E, are the i y after deposition and ly. As expected (Fig. the heating temperature. Accoltimg to Faraday’s law:

“‘dep

=-

if dep ZF

i being the current density, tdep the time of the aday’s constant and : =3. Sch zried [IS], assuming In di 111

rem

[email protected]“t,

(3)

Fig. 3. Infuence of the ~e~titl~ temperat on the relative intensity

of the in { 101)

revail over that of Sb, owing to the higher mobility of In, as shown by the ~\~uclr lower melting point: 150.61 and 630.7 1 “C for In and Sb, respectively. Note that in the case of both In and Sb diffusion, the obtained diffusion coefficient would represent the sum of the two indi~fidual coefficients. The diffusion coefficient depends on D * [2] according to:

synthesis of Mb at 100 “C. So far, their very high value obtained during the electrochemical synthesis remains unexplainable and perhaps not convincing. Indeed, by , >plying Schmalzried’s relationship [ 151 to their data, we found an even higher diffusion coefficient (lO_” instead of 10-‘h m” s-l).

3.3. Structural and pizysicochentical

and p re~reset~t gas constan molecular weight (236.57 .777 g cm- I), respectively.

(5)

e of the diffusion

Contrary to what was observed in the case of In electrodeposition on Bi cathodes, the In diffusion coefticient in InSb is very low, about five orders of magnitude lower than that in InBi where D is - 10 ” m’ s-’ at room temperature 131. This difference different structures. Most of the I: I compounds forme by elements of A and VA of the periodic table to which In, and belong (e.g., AIR AISb; CM, GaAs, GaSb; InP, InAs, InSb) crystallize with the zinc-blende (cubic) strucdrally coordinated atoms and strong g. 4(a), [ 1’31).The distance of the clo~st d Sb in IlnSb is for example 2.80 A. ptions m this series are the compounds i metallic elements, such as TlSb and TI the atoms are in the cesium-chloride (cubic) arrangement

the value ~nit~alyuand Sb were known, we evacuated from these moles formed on the unit surface area as a wcs.

value

As

aspects

for a cov~let~t

the

lower than

stance for the covalent bo

as a ~WVmeltin a more compact atomic ected to be more ‘impervi It is now worth recalling that in our case the

ones.

240

rtrongest

In iine as a function

of time after In deposition

and of the annealing temperature. In general, diffusion and reaction

density

lattice

defects)

( 1995)

in the solid

L. Peraldo Bicelli. G. Serravalle, J.

23.

S. Canegallo, V. Agripemo.

processes

C. Moruitou. A. Tou$\imi,

Bicelli. G. Serravalle. J. Alloys Comp. 234 (19%)

owing

to the

J.C. Wooley, J. Warner. Can. J. Phys. 42 (1964)

cover-

Y.N. Sadana, J.P. Singb. R. Kumar. Surf. Technol.

24

( 1985)

64.

so that only

the order of

magnitude of the average diffusion coefficient is significant. Sinre the geometry and structure af both the in deposit and the Sb substrate depend on the way tiley were prepared, the values we obtained for the Pn diffusion coefficient (e.g. D= 1O-‘0 m2 s-’ at 25 “C) refer to In electrodeposits on a bulk Sb substrate. A more detailed investigation on these aspects may be illuminating. Moreover, the strxtural and physicochemical considera’tions be shown the interest in investipsting the formation of other 1 to 1, HI-V compound5 and suggest promising research in this area, too.

Y.N. Sadana. J.P. Singh. Plat. Surf. Fin. 64 M.

Pcarbaix.

Atlas

of

Electrochemical

Solutions, 2nd ed.; NACE.

Houston, TX,

1879

I 1985)

Equilibria

in

319.

Aqlncous

and CEBELCOR,

Brus-

sels. 1974, p. 436 (In), 524 (Sb). I-.::.

hnassalski. Binary

Alloy

Society of Metals. Ohio.

Phase Diagrams,

Vol.2.

American

3986, p. 1395.

T. Okubo. M. Landau,

Abstract 376. The Electrochemical

Extended Abstracts,Vol.

88-2, Chicago. IL. Oct. 9-14.

J. Ortega. J. Herrero. J. Electrochem. G. Mengoli.

M.M.

Electrochem.

2I

M.C. (1965)

Musiani,

Hobson Jr., H. Leidheiser

136 (1989) M.

3388.

Gazzano,

J. Appl.

Jr., Trans.

Met. Sot.

AIME

233

482.

F.H. Eisen. C. Birchenall. ASTM

Sot.

F. Paolucci.

Society

1988, p. 547.

( I99 I 1 863.

Cards: S-0642.

Acta Met. S

( 1957)

1974 (In); 35-732.

265.

1990 !Sb): h-02%.

1967

( InSb). H. Schmalzried. Solid State Reactions. Academic Press. New York. 1974, pp. 53-t&

95-97:

D D. Wagman. V.H. SM.

124-127.

Evans. V.B. Parker. R.H.

Bailey. K.L. Chumey.

I I. Suppl. 2 (1982)

Schumm.

1. Halow.

R.L. Nuttnll, J. Phys. Chem. Ref. Darn

2-134.

C.A. Wert, R.M. Thomson,

C. Moraitou and A. Toussimi are indebted to the Time (Erasmus) program for a grant. The research was carried out with the financial su!>port of the I’*/linistero della ;‘niversitE e Jel!a Ricerca Scientifica e Tecnologica.

L. Peraldo

21 I.

(surface

morphology) and structure (single crystals or p;efened orientation., grain size, dislocation

and other

Demeneopoulos.

Alloys Comp. ?28

re~rodu~ibk phenomena state al-2 scarcely consistent influence of the crystal geometry age. material polycrystals,

S. Canepallo.V.

Physics of Solids. McGraw-Hill.

New

York, 1970. pp. 54-67. Landolt-Boemstein, Bitnd. Atom-

Zuhlenwerte

und Funktioncn. Sechste Aufliige.

und Molekularphysik.

4. Teil.

Ktistitlle,

Verlag. Berlin. 1955. p. 24 (B-2 type); pp. 24-25

t B- IO

type).

P Binnie. Acta Cryst. 9 F?Y. Chevalier. Cnlphad

( 19%) 6X6. ( I9XH) !t(3.

I2

I

Springer-

(B-3 type); p. 27