Applying nanomanipulation to the EBSD analysis of a gold wire

Introduction

As the trend towards miniaturisation and nanotechnology increases in microelectronic devices, a key growing requirement is the microstructural understanding of materials at the nanoscales to produce reliable products.

Detailed failure analysis on this scale typically requires a focused ion beam (FIB-SEM) combined with nanomanipulator tools to extract and lift-out specific regions of interest or failure. Analytical and structural characterisation of the lift outs is then usually undertaken in a TEM. This analysis is complex and there are limitations in the size of area which can be examined. Alternatively, the technique of electron backscatter diffraction (EBSD) can be applied for these applications, either in the FIB-SEM or the FEG-SEM fitted with EBSD systems.

This application note describes a method of combining Oxford Instruments' OmniProbe tools and AZtec EBSD system for the manipulation and analysis of a 5 µm diameter gold microelectronic wire sample.

Gold Wires

Gold wires in a range of sizes are typically used to make junctions on micro-electronics devices. Both the crystallographic texture and grain size distribution in these wires impact electrical and lifetime properties of the devices in which they are used. It is therefore critical to understand and control these parameters to improve both the performance and the reliability of these devices.

Owing to the fine size of these gold wires, conventional metallographic techniques for sectioning, mounting and polishing cannot be used. A solution at these nanoscales is to use an OmniProbe nanomanipulator combined with FIB-SEM tools to section regions of interest, to lift them out, and then to position them correctly for EBSD analyses. As an example, a 5 µm diameter gold wire is used to demonstrate this technique.

Experimental Procedure

Application of OmniProbe Tools for sample preparation, cutting and transfer procedure:

A FIB-SEM and the OmniProbe 100 was used to cut and transfer the sample to an SEM with an EBSD system. Figures 1 and 2 illustrate the series of steps undertaken to cut the wire and mounting and positioning for EBSD analyses. This is achieved through 5 steps:

1. Attach 5 µm gold wire onto carbon adhesive disc – Figure 1a
2. Navigate to region of interest and deposit platinum protective layer onto the wire – Figure 1b
3. Cut coated section using FIB – Figure 1c; Cut excess carbon adhesive disc from section – Figure 1d
4. Attach cut section to a lift-out OmniProbe needle – Figure 2a and 2b
5. The cut section is orientated and attached to a sample holder, in this case a standard TEM grid – Figure 2c and 2d

A

a) Identify area of interest

B

b) Deposit protective Pt layer with FIB.

C

D

c) and d) Cut for lift-out using OmniProbe 100.

Fig.1. Ion images: steps of sample preparation using FIB and OmniProbe Tools.

A

B

a) and b) Sample is removed and transported from gold wire to the analytical holder.

C

D

c) and d) Sample is reoriented for EBSD analyses and welded to a holder (a standard TEM grid) using Carbon deposition.

Fig.2. Attach and orientate section for analysis

EBSD Analyses

The sample is orientated and fixed to the TEM grid, it is then transferred to a FEG-SEM for EBSD analyses. The Oxford Instruments NordlysNano EBSD detector coupled with an AZtec system was employed to characterise the microstructure of the sample.

Fig. 3 shows a secondary electron image of the extracted gold wire section. The sample was tilted at 70° in a FEG-SEM. EBSD analyses was carried out using the parameters listed in Table 1.

Table 1: EBSD acquisition parameters details.

Fig. 3. SE image of gold wire section tilted at 70° for EBSD analyses. The grain structure can be clearly seen.

Results and Discussion

EBSD band contrast and IPF coloured maps are shown in Fig. 4. a and b respectively. It is clear from Fig. 4b that the crystallographic direction of the grains are strongly aligned with long the axis of gold wire. The pole figures in Fig. 6 show that this alignment is along the <111> and <001> directions and the wire has a dominant {111} fibre texture and a weaker {001} texture component. The fractions of the {111} and {100} components are 84% and 16% respectively. The latter is distributed mainly in the central region of the wire and a large grain to the top right of the section.

The average grain size in this section is 0.24 μm, a histogram of grain size distribution is shown Fig. 5. The relatively coarse grain observed on the right side of the section has a {100} orientation, is over 1.8 µm in diameter and dominates the cube component in pole figures in Fig.6.

Fig. 4. a) EBSD band contrast image.

b) IPF (Z direction) of the long axis of gold wire.

Fig. 5. Histogram of grain size distribution - average grain size in this section is 0.24 μm.

Fig. 6. Set of {100}, {110} and {111} pole figures from analysed gold wire sample.

Figure 7 illustrates a set of ODF sections at f2 (Euler3), showing a dominated {111} fiber texture of the gold wire and a cube {100} texture from the coarse grain. In order to quantify the preferred orientation tendency f(g), figure 8 shows the orientation density across the cube {100} and {111} components (for f1 and f2 = 45°) as a function of f. The maximum intensities found in the ODF are {111} and {100}.

Fig. 7. ODF sections at f2 showing the strong {111} fiber texture of the gold wire and a cube {100} texture from the coarse grain.

Fig. 8. Orientation density f( g) of gold wire across the cube {100} and {111} components for f1 and f 2 = 45° .

Conclusions

The analytical approach presented in this application note describes the accurate and straightforward way to prepare, manipulate and analyse very localised sample regions.

Acknowledgements

Oxford Instruments thanks Patrick Nguyen, Jean-Michel Desmarres, Florie Mialhe (CNES, France), Cedric Seguineau (FIALAB, France) and Helene Chauvin (THALES TCS, France) for providing the EBSD data, images and assistance.