X-ray scattering from interface dislocations in highly mismatched oxide epitaxial films


R. Guinebretière, A. Boulle, O. Masson, A. Dauger

 

Science des Procédés Céramiques et de Traitements de Surface, umr 6638. ENSCI, 47 Av. A. Thomas 87065 Limoges, France.

 

It is well known that the real physical properties of thin films are often greatly influenced by the microstructural characteristics of the film. Among microstructural parameters the state of strain of the film is probably one of the most important. During the last decade a large amount of work has been devoted to the determination of strains and strain gradients in epitaxial thin films. Nevertheless, most of the studies are related to semiconductor or metallic thin films. Since several years we are developing specific x-ray diffraction setup and modeling tools dedicated to the study of nanostructured oxide materials [1, 2]. The main specificity of those materials is that they exhibit high defect densities as compared to semiconductor materials.

In epitaxial thin films strains are generally related to the lattice mismatch across the film – substrate interface. The strain distribution is therefore highly anisotropic and moreover non constant along the normal to the interface. Recently, we proposed a new approach of the description of the strain gradient in epitaxial thin films [3]. This approach, based on the use of B-spline functions, allows one to obtain a precise description of the strain evolution along the normal to the interface. Experiments were performed on 60 nm thick epitaxial zirconia thin films deposited on sapphire substrate. As a result, we observed a strongly diverging behavior occurring near the interface in the first 5 nm of the film [3]. In comparison with the work done on semiconductor thin films we could imagine that such behavior is related to the presence of misfit dislocations. In this communication we give further evidence of the presence of such dislocations and discuss the possible effects on the XRD profiles.

The first step in such a study is to choose a convenient sample. Previous TEM studies on ZrO2 thin films grown on MgO showed that this system exhibits a simple cube on cube epitaxy [4,5]. The samples were elaborated through a specific sol-gel process already described elsewhere [4]. A thermal treatment of the samples at 1500°C during one hour induces the formation of zirconia islands epitaxied onto the magnesia substrate. The epitaxial relationships between zirconia island and magnesia substrate where confirmed through x-ray diffraction q-2q scans and f-scan. Those relationships are the followings:

(100)ZrO2 // (100)MgO and [001]ZrO2 // [001]MgO

 

The structure of the interface as been studied by high resolution x-ray diffraction. Reciprocal space maps (RSM) were measured using a specific x-ray diffraction setup already described in details [1,2]. The two dimensional intensity distributions were obtained in a one step measurement procedure. In spite of a large broadening of the reciprocal lattice points (RLP), those measurements were performed in only few hours. For instance, a RSM including the (200) magnesia RLP and the (200) zirconia RLP is given figure 1. According to the well known orientation conventions, Qz is perpendicular to the interface and Qx is lying in the interface. We recorded such maps at different values of the f angle and for two orders of the (h00) zirconia planes.

The mean thickness of the islands was determined by line profile analysis performed on qz sections of the (200) and (400) zirconia RSM. This mean thickness is close to t=100nm and the root mean squared strain is 0.044%.

 

Figure 1: Reciprocal space map including the (200)  zirconia RLP and the (200) magnesia RLP.

 

 

Figure 2: transverse scans of the (h00) reciprocal nodes. (a) qx scan across the (200) ; (b) qx scan across the (400).

 

In the MgO/ZrO2 system the theoretical lattice mismatch is close to 20% and strain relaxation is therefore highly expected in particular through the generation of misfit dislocations. The presence of dislocations lying at the interface must induce modification of the scattering profile in the Qx direction. A section along this axis, of the (200) zirconia RLP is given figure 2a. This scan exhibit a peculiar line shape made of two symmetrical shoulders around the main peak. It should be noticed that such line shapes have already been observed by other authors in sample containing misfit dislocations [6, 7].

The understanding of the physical meaning of the presence of those satellites requires the comparison of several scans obtained for several reflection orders. The transverse scan of the (400) zirconia RLP is reported figure 2b. The two satellites are also clearly visible. An interesting feature is that both profiles are perfectly superimposed (after rescaling by a constant factor) on an angular scale. This  demonstrates that the effects responsible for this peculiar line shape are linearly dependent of the order of reflection and no size effect is observed. Such behavior is characteristic of a disorder that is rotational in nature [8]. In particular, the addition (or substraction) of half planes not perpendicular to the interface (e.g. with a non zero z-component of the Burgers vector) can explain such a behavior. The presence of the two secondary maxima could be related the existence of two equivalent glide planes for those dislocations.

Further details concerning the XRD analysis as well as the possible slip systems will be given at the conference.

 

1. A.Boulle, O.Masson, R.Guinebretière, A.Lecomte and A.Dauger. J.Appl.Cryst. 35(2002)606-614.

2. A.Boulle, O.Masson, R. Guinebretière and A.Dauger. Diffraction analysis of the microstruture of materials. Springer Series in Materials Science, vol. 68, 2004, ISBN 3-540-40519-4 Pages 505-526.

3. A. Boulle, O. Masson, R. Guinebretière, A. Dauger. J.Appl.Cryst. 36(2003)1424-1431.

4. R.Guinebretière, B.Soulestin and A.Dauger.  Thin Sol. Films. 319(1998)197-201.

5. R.Guinebretière, A.Dauger O.Masson and B.Soulestin. Phil.Mag. A 79[7](1999)1517-1531.

6. V.M.Kaganer, R.Köhler, M.Schmidbauer, R.Optiz and B.Jenichen., Phys.Rev. B 55[3](1997)1793-1810.

7. A. Yu Babkevich, R.A.Cowley, N.J.Mason, S.Weller and A.Stunault. J.Phys.Cond.Matter 14(2002)13505-13528.

8. P.F.Miceli and C.J. Palmstrom. Phys.Rev. B 51[8](1995)5506-5509.