MEIS analysis at the nanoscale, advantages/disadvantages versus X-Rays and TEM

1. Univ. Grenoble Alpes, INAC-SP2M, LEMMA, F-38000 Grenoble, France
2. CEA, INAC-SP2M, LEMMA, F-38000 Grenoble, France

Numerous analyzing methods have been successfully and widely employed in material sciences in order to obtain chemical, elemental, morphological and structural information in bulk materials as well as in layers of various thicknesses. The emergence, since the middle of the 80s, of nanoscience poses new questions for analysis techniques, including IBA, in terms of spatial resolution and sensitivity. The challenge posed by the analysis of nanostructures is different in each case. Rather than attempting to describe all the possible situations, I will focus on the structural study of single crystal nanostructures.

GaN quantum dots deposited on (and embedded in) AlN give an interesting example of structural analysis at the nanoscale. Indeed, the strain state of these quantum dots embedded in a matrix is an important parameter which governs to some extent their optical properties. The most usual wurtzite crystallographic phase of these two materials exhibits both piezoelectric and spontaneous polarization. Because of the elevated value of the piezoelectric constants, a huge internal electric field is currently observed in nitride heterostructures. As a consequence of the resulting quantum confined Stark effect, a strong red shift is induced, leading to luminescent emission at energies smaller than the GaN gap value for dots higher than 2.5 nm.

Amongst the different available techniques to measure strain, the Geometrical Phase Analysis (GPA) of HRTEM images, has the advantage to provide strain map of the nanoobject with, in principle, an atomic resolution [1]. However, large strain fluctuations are observed in the strain map along the c axis and, in addition, the expected vertical strain gradient is not visible on these images. The comparison with the electron diffraction techniques, namely CBED and NBED, will also be discussed.

An alternative possibility is the use of x-ray diffraction, in grazing incidence to enhance the contribution of the dots (and AlN capping) with respect to that of the substrate. More precisely, by varying the scattering power of a selected element, for instance Ga, it is possible to localize specifically GaN regions in reciprocal space. This is the purpose of Multiwavelength Anomalous Diffraction (MAD) measurements, which consists in recording the scattered intensity as a function of the energy across the absorption edge of an element, namely Ga in the present case. The position of the Ga signal maximum along the [10-10] direction is directly related to the average in-plane strain state and structure in the QDs [2]. The diffraction peak corresponding to the GaN QDs is clearly due to the presence of an in-plane strain gradient within the dots but also to the finite lateral size of the dots (about 40nm). Therefore, the extraction of a strain profile is not straightforward and model dependent.

By contrast, Medium Energy Ion Scattering (MEIS) can measure the deformation profile of quantum dots with a depth resolution in the monolayer range [3]. Indeed, ion blocking reveals the angular position of a nucleus with respect to the scattering center that can be directly related to the out-of-plane to in-plane lattice parameters ratio. Of course, the ion beam spot size is much larger than the dots and the MEIS measurement averages the strain profile over about 6 × 108 dots. From this point of view, MEIS is similar to X-rays techniques but the contrary to X-rays, the depth strain profile along depth within the dots can be accurately measure and its dependence clearly appears comparing the profile of uncapped dots in figure 1(a) to the one of capped dots in figure 1(c). Another example of a MEIS structural analysis will be given by the measurement of in-plane and out-of-plane mosaïcity of a set of GaN nanowires and the comparison with XRD results.

[1] E. Sarigiannidou et al., Appl. Phys. Lett. 87 (2005) 203112
[2] J. Coraux et al., Phys. Rev. B 74 (2006) 195302
[3] D. Jalabert et al., Phys. Rev. B 72 (2005) 115301