Mechanical Strength 

Get Complete Project Material File(s) Now! »

PART 2: MECHANICAL STRENGTH EXPERIMENTAL

In situ SEM tensile tests on nanoporous gold have not been done, likely due to the technical difficulty of testing such a delicate and macroscopically brittle material, and being able to capture the deformation in the correctly predicted area. Since our tests involve a thin foil, special precautions must be made to avoid unintended fracture. For this reason, a special tensile test sample was designed to ensure fracture would occur in the correct specified location to ensure that a fracture could be captured under an SEM.
Tensile test specimens were made by cutting a copper sheet of 100 μm thickness, into 3 x 0.5 cm. Using an X-Acto knife, a single thin approximately 20 μm in thickness line was scratched onto the surface. Samples of 3 x 10 mm nanoporous gold were placed over the line so that a portion of the sample was free standing over the cut. The ends of each sample were glued with super glue, to secure it in place. A diagram of this design is shown in Figure 3.1-1.
Tensile tests were performed using a 5000 N Tensile/Compression Module at a speed of 0.5 μm/sec with a 50 newton load cell. A serrated vice grip was used to secure each ends of the copper sheet during the test, so that the nanoporous gold remained untouched by the grips. With this design only the glue and friction between the copper and nanoporous gold surfaces caused tensile and shearing stress on the material. Videos were made by capturing high speed frames through the SEM and putting them together. Preliminary tests were also performed under an optical microscope. Videos were captured with a digital camera through the optical microscope.

RESULTS

Preliminary tests were done outside of the SEM to ensure this technique would work in the SEM. The important results that we were looking for was that, one, the nanoporous gold would not sink into or break from its own weight over the slit in the copper foil sample, and two, the nanoporous gold would break at the predicted location over the slit of the copper foil. The experimental technique proved to be successful. SEM inspection of the samples before deformation showed that nanoporous gold was successfully free standing over the slit that was cut in the copper foil, and there were no significant fractures before the tensile test. This can be seen in Figure 3.2-1, where a tensile test sample was image in an SEM.
Using this specimen, the tensile test was then performed outside of an SEM with an optical microscope. A video was captured of the fracturing occurring. The video showed that during deformation multiple cracks propagated throughout the entire sample, due to the copper deforming underneath and putting stress on the nanoporous gold. However, a crack occurred where we predicted it would, making this technique useful to use in the SEM.
After this proved successful samples were tested in an environmental SEM. The tensile test was performed at a speed of 0.5 micrometers per second. Several iterations were done before data was successfully captured. The optimal way to perform the tensile test and capture deformation of nanoporous gold was to first observe the tensile deformation of the sample at low magnification. Once a crack was observed in the copper sample, the tensile test was paused, and magnification was increased until micro cracks could be observed on the nanoporous gold. A microcrack was then chosen, and magnification was increased until the ligament morphology was visible. The ESEM was set up so that an image was saved every 0.5 seconds. The tensile test was resumed and deformation of the crack was observed. When significant changes were noticed in the crack, the tensile test was paused, and the focus and stigmation were adjusted, and low speed images were captured. The tensile test was resumed, and this process repeated several times, until either the crack stopped propagating or the image quality become too poor to capture useful data.
The propagation of cracks in nanoporous gold are shown in Figure 3.2-2 through Figure 3.2-4. The first two figures show how a crack interacts with a solid gold grain boundary. In both figures green circles locate the area where the crack goes through the grain boundary. In both these locations the crack propagates underneath the grain boundary, initially leaving the grain boundary intact. The grain boundary however becomes deformed and starts to neck in the center. In Figure 3.2-2, the crack goes through the grain boundary on the left side in image d, but this time grain boundary does not remain intact, and instead almost immediately fractures, an inconsistency in the interactions between crack propagation and grain boundaries.
Significant pore expansion, shown in the yellow circles, occurs next the grain boundary as the crack passes through. Along the grain boundary there are often locations that have a high density of porous space, which may be a weak location for the crack to follow through to. This was noticed in Figure 3.2-3, where the fracture began to follow the edge of the grain boundary, before going underneath it.
Cracks tend to follow the path that contains smaller ligaments and larger pores. These smaller ligaments tend to stretch and pinch off more easily than larger ligaments, and the larger pores tend to expand faster than smaller pores. There are several examples of these occurrences throughout all the images. In Figure 3.2-3 an example of a pore expanding next to a grain boundary is shown in the yellow circle. In the first image, this pore is noticeably larger than the pores surrounding it. In the next image this pore has become significantly larger and more circular. The ligaments around it have stretched and become thinner, possibly indicating plastic deformation. However, in the last image we see that ligaments around the pore are barely visible, meaning that they have curled on itself or retracted, indicating that there may still be some elasticity in the ligament. On the upper right of this same figure, an outline of a pore and its developments as deformation continues is shown. In this example, the pore is also next to a grain boundary. The grain boundary seems to provide a stable area, blocking the pore from expanding in that direction.
Figure 3.2-4 also shows several examples of pores expanding and ligaments pinching off. In these images we found a location where a crack was about to form, but had not opened up yet. It was noted early on that before cracks form or propagate through an area, this area tends to be darker than the rest of the area, something caused by possibly pore expansion, thinner sample area, or deformation already occurring underneath the surface of the sample. For this reason, this area was chosen to observe. Using a reference point (the small speck below the fracture), several colored lines in the images point to different features near the crack tip, and helps you follow their shape changes, as the sample deforms. The blue line to the very left shows a thin ligament with about seven other smaller ligament attached around it. As the crack begins to form, the pores around this ligament start to expand, and the connecting ligaments stretch and become thinner. Eventually, at image d, most of the thinner ligaments have detach themselves from this structure and only two points, on both fracture surfaces, hold the structure together. The last image shows this ligament structure after the fracture has occurred, and it has become very thin and elongated, almost to the point of no longer being visible. The yellow line to the right of the blue line, points to a porous structure in a roughly ‘M’ shape, with a very this ligament on its left side and more pores on its right side. As the crack propagates this structure begins to expand, the thin ligament breaks, and this space opens up to become the two fracture surfaces. Finally, the red line shows a pore, above which a fracture almost occurred, but not quite. During the progression of the images, there is a slight pore expansion up to image d, caused by the applied stress on each side, to the point where a full fracture has almost occurred. However, the fracture occurs further above this location, after which, pore has shrunk in size. This may indicate that there is some elasticity after the fracture. There are other examples of pore expansion and then relaxation after the crack passes through. In Figure 3.2-2, and yellow circle shows a pore that begins to expand the closer the crack gets to it. However, after the crack passes through the pore relaxes, going back to its original size.
An important observation in general after this fracture, is that there are several ligament structures hanging off the fracture surface which have been bent and deformed to the point that they are almost unrecognizable when comparing them to the previous image. This indicates that there is some plastic deformation During crack propagation, more often than not, two or more cracks form in close proximity, an example of which can be seen in Figure 3.2-2 a and b. The two fracture points are labeled 1 and 2. Fracture point 2 forms above fracture 1, however it stops propagating once point 1 joins with the left end of the 2^nd fracture. This leads to large chunks of ligament structures hanging off the fracture surface, shown in red circles in Figure 3.2-2 c and d.
An important observation, especially based on these images, is that after fracture occurs, the fracture surface tends to whiten in the SEM images. This may indicate that ligaments are flipping up toward the electron detector, or that ligaments are curling in on themselves. This may be something that occurs immediately after fracture, or takes some time to develop.
In summary we noticed several things occurring during crack propagation: 1. Grain boundaries can have varying effects on crack propagation, the reason for which needs further discussion. 2. Pores expand, in general into more circular shapes. 3. Ligaments pinch off, rotate, and deform significantly, as can be seen on a fracture surface. 4. A dark area forms before a crack passes through. 5. Edges of fracture surface sometimes turn white on SEM images. 6. Cracks propagate through locations with smaller ligaments and larger pores.

DISCUSSION

The first observation noted in the results section was that grain boundaries had varying effects on the crack propagation, in some cases assisting and in others impeding the crack propagation. In Figure 3.2-2, there are two locations on the grain boundary where the crack propagated through. In one case the grain boundary is strong and remains intact while the crack goes through it (right) and in the other the grain boundary fractures almost immediately (left). There are a couple of possible reasons for these differences. First, it is observed that the location that the crack occurred is a location where the grain boundary is much thinner, making fracture of this structure easier. Second, the angle of this side of the grain boundary to the direction of the crack propagation could have an effect on the crack propagation. It is fortunate in our studies that we found cracks propagating through grains at three different angles. In this case, the crack to the left goes through the grain boundary at almost 90 degrees, much higher than the angle of the right side grain boundary to the direction of the crack, which is about 45 degrees. In Figure 3.2-3 the crack actually follows the edge of the grain boundary at 0 degrees, before changing its direction and going through the grain boundary. All of these angles seem to have a different impact on the crack propagation. In the case of 90 degree angle, the grain boundary nearly immediately fractured, impeding the crack propagation very little. In the 45 degree angle case, the grain boundary remains firmly intact, slowing down the propagation of the crack. This may be due to the distance across the grain boundary relative to the direction of the crack, is higher than it would be if the grain boundary was at 90 degrees relative to the crack. Finally when the crack is 0 degrees to the grain boundary, in seems to help the propagation of the crack rather than impede it. Since there are a high density of pores next to the grain boundary, the crack is able to propagate through this porous area.
Throughout Figure 3.2-2 through Figure 3.2-4 we see several examples of possible elastic and plastic deformation. The examples of elastic deformation are as follows. We have seen that pores expand when a crack comes close to it, and then relax back into its original shape when the crack passes by the pore (but not through it). At the same time ligaments next to these pores stretch, decreasing their diameter initially, and then increasing after the crack passes through. Examples of plastic deformation are shown at the crack surfaces where ligaments are stretched and deformed significantly, and they remain that way after the crack passes through. Also on the fractures surfaces, whitening has been noted on the ligament. This could indicate that in these cases ligaments have risen toward the detector, increasing the signal. This could also indicate charging in this area. Since deformation in metals in some cases, increases the resistance, this may also indicated plastic deformation.

TABLE OF CONTENTS
Abstract 
General Audience Abstract 
Acknowledgment 
Table of Figures 
1 Introduction 
2 Part 1: Morphology Tuning 
2.1 Experimental
2.2 Results
2.3 Discussion .
3 Part 2: Mechanical Strength 
3.1 Experimental
3.2 Results
3.3 Discussion
4 Conclusions 
5 Appendix 
5.1 AQUAMI: Morpholgy image analysis program
5.2 Wash Solution Experiment
6 References
GET THE COMPLETE PROJECT
Morphology Tuning and Mechanical Properties of Nanoporous Gold