A. CHARAI, D. ROUX, A. ROLLAND, G. SAINDRENAN* and B. AUFRAY. Laboratoire de MBtallurgie, CNRS-UA 443, Facult6 des Sciences et. Techniques de ...
JOURNAL DE PHYSIQUE Colloque C5, supplement au nolO, Tome 49, octobre 1988
INTERFACIAL STRUCTURE MODIFICATIONS INDUCED BY SULPHUR IN Ag-Ni ALLOYS A. CHARAI, D. ROUX, A. ROLLAND, G. SAINDRENAN* and B. AUFRAY
Laboratoire de MBtallurgie, CNRS-UA 443, Facult6 des Sciences et Techniques de Saint-JBr6me. Av. Escadrille Normandie-Niemen, Case 511, F-13397 Marseille Cedex 13, France "ENSM, 1, rue de la NoB, F-44072 Nantes Cedex, France
R6sum6 - L'influence de la segregation du soufre 5 l'interface de deux m6taux de faible solubilit6 mutuelle (Ag/Ni) est dtudi6e par microscopic dlectronique en transmission (M.E.T.). Les interfaces (100) sont obtenues par dvaporation successive d'argent et de nickel sur substrat monocristallin de NaCl clive. Les observations effectudes en M.E.T. montrent que le nickel est en bonne Gpitaxie sur l'argent alors que les dislocations de raccordement ne sont pas visibles. L'introduction de soufre provoque une decohdsion de cette interface. Abstract - In this paper we present the first results concerning the effect of XI iapurity of low solubility on the interfacial structure of simple model system : the Ag/Ni system with sulphur as the impurity common to the two metals. The Ag/Ni interfaces are produced by vapour deposition and epitaxial growth on single crystal substrate of NaCl cleaved. T.E.M. observations of these interfaces show that nickel is in good epitaxy on silver while misfit dislocations network is not visible. After annealing in pure hydrogen, epitaxy is reinforced : this indicates that the (100) interface is preserved. Quite the contrary, after annealing under H /H S ambient, the (100) 2 2 interface disappears.
I
-
INTRODUCTION
Interfacial composition between two phases may be very different to volumetric compositions of the two adjacent phases. This is due to the very localized nature of structural discontinuity and/or the properties of the two phases present. It is generally supposed that the level of segregation phenomena at interfaces depends on the chemical nature of the elements present and on the structure of the interface. Despite the technological importance of this phenomenon, there is little, if any, experimental data available concerning interface segregation / 1 / In this paper we present the first results concerning the effect of an impurity of low solubility on the interfacial structure of a simple model system : the Ag/Ni system with sulphur as the impurity common to the two metals. This choice of this system was dictated by two factors - firstly, the low mutual solubility of Ag and Ni (which limits transport of matter to the interface) ; and secondly, knowledge acquired in our laboratory, regarding the binary and ternary systems Ag(S), Ag(Ni) and Ag(Ni,S) from the point of view of superficial and intergranular segregation 12-51.
.
I1
-
EXPERIMENTAL PROCEDURE
An important point of our experimental procedure consists in the fact that both T.E.M. observations and thermal treatments are performed on the same thin Ag/Ni film. This experimental procedure includes six stages :
Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1988551
JOURNAL DE PHYSIQUE 1- Realisation of a thin film of Ag/Ni (bicrystal) 2- Interface observations by T.E.M. 3-Anneal under hydrogen flow at 500°C for 4 hours 4- Interface observations by T.E.M. 5- Introduction of sulphur into the thin film by a diffusion process at 500°C for 4 hours 6- T.E.M. observation of possible interface structural modification due to sulphur segregation.
The Ag/bli thin films are produced by vapour deposition and epitaxial growth on single crystal substrate of NaCl ((100) orientation). The main experimental condition are given in table below : Substrate temperature
Evaporation rate
Thickness film
300°C
0.01 nm.s-I
45 nm
30°C
-1 nm.s
15 nm
0.05
After dissolution in water of Na-C1, the thin films Ag/Ni are set on nickel grids. 11.2.1. Hydrogen treatment This treatment is performed under high purity hydrogen flow at 500°C during 4 hours. This anneal makes it possible to obtain thin film in thermodynamical equilibrium. 11.2.2. Sulphur deposit The elementary sulphur is brought from the vapour phase into thin films by a diffusion process. A H2/H2S mixture, the composition of which is known and remains constant, circulates above the sample for the whole period of treatment. The experimental device is described in /6/ (Fig. 1). Let us recall the main points : - In the furnace F1 is placed the sample for treatment ; - In the furnace F2 is set a powder of copper sulphide ; with H2 (150 Torrs) the temperature of this furnace makes it possible to fix the ratio of partial pressures p~ S/pH , in other words the sulphur chemical potential ; - ?he e$ectromagnetic pump ( P ) is used to circulate the H 2/H2S mixture from F2 to FI. By this process the sulphur chemical potential can be arranged in the vapour phase so that no 3D sulphide (Ag2S or Ni3S2) can form on thin films ; the range is determined from the stability of the different sulphides (Pig. 2) 171.
FIGURE 1 : Apparatus for heat treatment in H -H S mixtures. 2 2
Our experimental conditions are : * Thin film temperature : 500°C * Copper sulphide temperature : 480°C * Anneal time : 4 hours
.d AG'(KJ
FIGURE 2 Free energy diagram for 3D sulphides of metals.
I
400
11.3. T.E.M.
600
T l"C]
obne/rvatLo~
The specimen is introduced into a Philips E.M. 400 T microscope (Cs = 2.7 mm), operating at 100 Kv (wave length = 3.7 pm). The objective aperture suppressed spatial frequencies superior to 1.5 nm-1. The electron beam is parallel to [loo] silver axis, so eight multiple diffraction beams were used to form moire patterns 131. Note that all the samples were observed under these conditions 191. 111
-
RESULTS
Before annealing, several types of moire fringes are observed on electron micrographs in very small fields (Fig. 3a). These areas are slightly disorientated in relation to each other, as confirmed by the elongated shape of nickel diffraction spots (Fig. 3b).
FIGURE 3 : a) Electron image showing Ag/Ni sample in epitaxy following direction [loo] ; before heat treatment. b) Corresponding diffraction.
JOURNAL DE PHYSIQUE
-432
After annealing, under hydrogen, of this interface, the network of moir6 patterns is visible on much larger areas than previously observed (Fig. ha). The double diffraction spots (which are the cause of these moird patterns) are more intense and better defined (Fia. 4b).
FIGURE 4 : a) Electron image obtained after treatment in presence of hydrogen at 500°C for 4 hours. b) Corresponding diffraction.
After introduction of sulphur, the electron micrographs obtained show : almost complete disappearance of moird patterns (Fig. 5a) ; corresponding diffractions (Fig. 5b) show that only silver remains orientated according to direction [loo]
-
.
FIGURE 5 : a) Electron image obtained after introduction of sulphur at 500°C for 4 hours. b) Corresponding diffraction. In addition, two distinct types of area may be :
i) Widespread areas of silver (free of defects) which are orientated according to the original direction [I 001. ii) Disturbed areas where silver and nickel coexist with no epitaxial relationship (Fig. 6).
FIGURE 6 : a) Electron image obtained after introduction of sulphur at 500°C for 4 hours. b) Corresponding diffraction. IV
- DISCUSSION AND CONCLUSION
Modifications observed after anneal under hydrogen indicate a strengthening of epitaxy between nickel and silver. The Ag/Ni interface has not only been maintained, but has stabilized. This result may be interpreted as follows : in metalloids purification of the interface and/or modification of the interfacial structure linked to temperature. Given that this behaviour is observed on thin films (thickness = 60 nm), this result shows that the (100) Ag/Ni interface is of very low energy. In addition, no observations are noted of misfit dislocations, which, if they existed, could be confused with moire patterns. We attributed this, in an initial hypothesis, to the great difference in crystalline parameters between nickel and silver (15%), giving rise to a network of interfacial dislocations of low periodicity. As a second hypothesis, we may consider that there exists an interface without misfit dislocations, indeed, we observed, elsewhere /3/ that in Ag(Ni) solid solutions, Ni segregates with a very unusual structure : the uppermost layer consists purely of solvent and nickel gathers underneath. On the other hand the influence of sulphur is seen to be particularly opposed to the very existence of this interface. Indeed, an almost complete separation of the two metals is observed, even though the chemical potential of sulphur is greatly inferior to that necessary for the formation of 3D sulphurs (Ni3S2 and Ag2S). This result may be explained by strong S-Ni cosegregation on the surface. This cosegregation, observed elsewhere / 4 / , would greatly reduce superficial tension in silver, thus destabilising the Ag/Ni interface. In order to better understand the origin of this drastic effect of sulphur on interface it would be preferable to determine the initial atomic structure of Ag/Ni interface. This envisaged, using H.R.E.M. observations of this interface on the cross-section.
JOURNAL DE PHYSIQUE
C5-434
Acknowledgment - We should l i k e t o thank P r o f e s s o r J.P. PETRAKIAN, f o r t h e u s e of h i s e v a p o r a t o r system and Dr. A. BOURRET f o r h e l p f u l s u g g e s t i o n s and advice.
References JOHNSON and J.M.
W.C.
J. WANE
/3/
B. AUFRAY, F. CABANE-BROUTY, J. CABANE
-
BLAKCLY
-
/I/ /2/
J. de Chim. Phys.,
-
I n t e r f a c i a l Segregation 83, n o 11-12,
-
-
Eds. ASM, (1979)
709, (1986)
Acta Met.,
27, 1849-1854,
/4/
A. ROLLAND, B. AUFRAY
/5/
B. AUFRAY, A. ROLLAND, P. GAS
/6/
J. OUDAR
/7/
J. BENARD, J. OUDAR, N. BARBOUTH, E. MARGOT and Y. BERTHIER 88, L35-L41, (1979)
/8/
C.A.
/9/
A. CHARAI, D. ROUX, A. ROLLAND, B. AUFRAY 29a (1987).
-
S u r f a c e Sci.,
MCtaux, Corrosion
BASSET, J.W. MENTER, D.W.
-
162, 530-537,
Surface Sci.,
(1979)
(1985)
178, 872-879,
(1986)
I n d u s t r i e s , 423, 397, (1960)
PASHLEY
-
-
Surface Sci.,
Proc. Roy. Soc. A 246, 345, (1958)
-
J. Microsc. Spectro. Electron.,
