HAL Id: jpa-00220897
https://hal.archives-ouvertes.fr/jpa-00220897
Submitted on 1 Jan 1981
HAL
is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire
HAL, estdestinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
DOMAIN MICROSCOPY IN CHALCOGENIDE ALLOY GLASS THIN FILMS
J. Phillips
To cite this version:
J. Phillips. DOMAIN MICROSCOPY IN CHALCOGENIDE ALLOY GLASS THIN FILMS. Journal
de Physique Colloques, 1981, 42 (C4), pp.C4-197-C4-200. �10.1051/jphyscol:1981440�. �jpa-00220897�
JOURNAL DE PHYSIQUE
CoZZoque C4, suppZQrnent au nO1O, Tome 42, octobre 1981 page C4-197
DOMAIN MICROSCOPY I N CHALCOGENIDE ALLOY GLASS T H I N F I L M S
J.C. Phillips
BeZZ Laboratories, Murray H i Z Z , N.J. 07974, U.S.A.
Abstract.- We report the observation of very well resolved polygonal domain (not island) networks in thin glass films. The domains are observed by electron micros- copy and their hexagonal diameters are as large as 1000 ( + 100)A. The composi- tion dependence of the average diameters of the polygonal structures correlates well with the glass-forming tendency and agrees very well with the predictions of a recent topological model.
The morphology of non-crystalline solids often varies dramatically depending on the method of preparation, while the short-range order (and properties such as the paramagnetic spin density, which depend primarily on the s.r.0.) is much less variable. The morphology of a-Si, for example, is quite different if it is prepared by evaporation or by ion-bombardment of previously crystalline material,' yet there are no noticeable differences in electronic pro- perties in the two materiak2 On the other hand, it has generally been supposed that the morphology of a congruently melting and subliming very good glass former (such as As2Se,) would be much the same whether it was quenched from the melt or evaporated as a thin film. It has also been generally assumed that evaporated films made from good glass formers would be continuous and would not display the columnar or granular structures which are generally observed in films prepared from amorphous materials which upon heating in bulk are found to crystallize before undergoing a glass transition.
These assumptions are derived from the general idea that network glasses are properly described structurally as fully relaxed and strain-free continuous "random" networks. The validity of this viewpoint has recently been questioned on theoretical ground^.^ It has been pointed out that in most non-crystalline network solids (such as a-Si) strain energy accumu- lates in the network as a cluster of atoms grows- larger. The discontinuous structures observed in a-Si are a consequence of this accumulated strain, which can also be interpreted kinetically in terms of island non-coalescence or non-intergrowth.4 Even in network glasses Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1981440
JOURNAL DE PHYSIQlTE
in which the strain energy and configurational entropy are minimized, however, these effects are still present, but to a much lesser degree. As a result the granular structure should still be present, but with a much reduced density, i.e., on a much larger scale. This means that morphological instability is expected even in the most nearly "perfect" glass. which is3 As2S3 or As2Se3.
We have recently observed5 this instability by transmission electron microscopy in a series of evaporated As2-, (Sel-yTe,)3+x films based on As2Se3. The general features observed in the films are: the optimun thickness is 500A. The optimum composition is x = y = 0. At this composition the usual island morphology is replaced by a domain mor- phology. The domains are about IOOOA in diameter on the average. They are separated by deep, extremely linear troughs about 30-50A wide. With y = 0 increasing x degrades the troughs rapidly without changing the domain diameters, and the degradation to an island-like morphology is nearly symmtrical in f x and almost complete at 1x1 = 0.1. With x = 0 increasing y gradually reduces the domain diameters from IOOOA (x=y=O) to lOOA (x=O, y=l). The morphology of the latter limit is typically columnar, as expected for a-As2Te3.
The best evaporated Si:H films that we have seen resemble As2SezTe both in domain diame- ter (--300A) and trough irregularity. Thus the passivating effect of hydrogenation on a-Si at best makes it morphologically similar to a poor glass, and one can regard the alloy series As2 (Sel-,Te,)3 as a system which spans the homogeneous glass-+homogeneous amorphous material transition. The degradation of the troughs in As2+,Se3-, for 1x1 0.1 is suggestive of morphological phase separation which is analogous to formation of (SiH2), regions in SiH, for x 2 0.3.
To emphasize the distinction between island formation and domain formation we show on the next page a TEM photograph of a 5OOA As2Se3 film? The domain structure is quali- tatively different from the island structure which is normally observed on thin amorphous films. This reflects the "nearly perfect" nature3 of the g-As2Se3 covalent network.
References
1. FOTI, G., BEAN, J. C., POATE, J. M., and MAGEE, C. W., Appl. Phys. Lett. 36 (1980) 840.
2. THOMAS, P. A., BRODSKY, M. H., KAPLAN, D. and LEPINE, D., Phys. Rev. B18 (1978) 3059.
3. PHILLIPS, J. C . , Phys. Rev. Lett. 42 (1979) 1151; J. Non-Cryst. Solids 34 (1979) 153;
Phys. Stat. Sol. (b) 101 (1980) 473.
JOURNAL DE PHYSIQUE
4. PHILLIPS, J. C., J. Non-Cryst. Solids 43 (1981) 37.
5. CHEN, C. H., PHILLIPS, J. C., TAI, K. L., and BRIDENBAUGH, P. M., Solid State Comm. (1981, in press).
