Scratch Testing of Metallic Nanofilms for Flexible Electronics

Metallic nanofilms on polymer substrates are extensively used as interconnect components in various flexible electronic devices, such as antennae structures, identification tags, solar cells, electronic skin, wearables and paper-like electronic display.^1,2 This is because of the inherent benefits of metallic nanofilms such as low specific weight, mechanical flexibility, ease of integration and low cost, as compared to silicon-based nanofilms.

The polymer substrate mostly endures the service loads, while the metallic nanofilm retains its function without rupturing until a relatively large amount of strain is applied. The metallic nanofilm has excellent electrical and mechanical properties, especially in electrical conductivity, yield strength and fracture toughness under the action of cyclic straining.

However, performance issues are often faced in overall reliability and durability of the flexible device and therefore, the adhesion property of a nanofilm to its substrate is crucial to prevent such issues. A poor adhesion property can contribute to a premature failure of the film and thus render the flexible electronic components ineffective. Fractures and interface delamination are the two most common failure modes of such nanofilms.

Being a reliable and relatively fast method, scratch testing can be easily implemented for inline quality control of films during production.^3-5 Nevertheless, scratch tests cannot be easily performed on such flexible and thin films. The use of a nanomechanical tester for such evaluation is limited by the force range to initiate failure of the film during scratch, as their sharp tips can cut through the film without cracking and delaminating.

It is the same with nanomechanical tools, which have limited productivity because of time-consuming test preparation. One system that is specifically designed to address this issue is the Universal Scratch Test system from Bruker. The system performs faster, easier and more reliable critical load scratch failure evaluations of nanofilms.

The UMT TriboLab Scratch Test System

Built on the UMT TriboLab™ platform, the scratch test system provides precision control of position, load and speed. The modular design of the system ensures the flexibility to scratch test across different velocities and forces. The UMT TriboLab Scratch Test System utilizes three vital drive systems — slider, Y-stage and carriage for X-, Y- and Z-motion, respectively. The tester is a highly versatile, user-friendly and highly productive scratch tool, thanks to integrated “intelligent” hardware (TriboID™) and software (TriboScript™) interfaces.

The TriboID feature automatically detects the assortment of components connected to the system and even configures them. TriboScript provides an improved and secured scripting interface for effortless compilation of scratch test sequences from the integrated test blocks. In addition, the system is integrated with real-time control and data analysis software to guarantee high repeatability and accuracy.

Bruker’s scratch test system can be effectively used for all modes of scratch testing, including constant and progressive load conditions. The carriage drive system provides the motion along Z-direction for displacement and loading. It also accommodates the slider drive that houses an optical microscope and a force sensor. The slider offers motion along the X-direction. The test specimen is mounted and movement is provided along the Y-direction by using the linear stage.

The scratch test system includes options for the simultaneous measurement of electrical surface resistance (ESR), electrical contact resistance (ECR), acoustic emission (AE), optical microcopy for automated imaging of the entire scratch, and in-situ scratch depth profiling using a capacitance sensor for tip-displacement measurement. Using Bruker’s data viewer software, the image of the entire scratch can be plotted along with other data, such as normal force (Fz), lateral force (Fx), AE, scratch depth, scratch distance, ESR and ECR.

A number of force sensors [FL: 5 to 500 mN; FVL: 1 to 100 mN; DFH series: 0.5 to 200 N; DFM series: 0.05 to 20 N] and scratching styli/tips [diamond stylus: 2.5, 5, and 12.5 µm tip radius; Rockwell indenter: 200 µm tip radius with 120° cone angle; Knoop indenter with two apex angles (130° and 172°50/); Vickers indenter (4-sided pyramid with a 136° apex angle); microblade (diamond, tungsten carbide)] are available.

Test Method

A DFH-1 force sensor and a tungsten carbide ball (1.6 mm diameter) were used to perform the scratch test of the metallic nanofilm on a polymer substrate. The film specimen was mounted on the Y-stage, followed by installing the ball under the force sensor. The scratch test was carried out by applying an initial load of 0.2 N on the film with the ball, the specimen was then moved over a distance of 2 mm at a velocity of 0.02 mm/second.

The normal load (Fz) was linearly increased from 0.2 to 8 N during the movement of the test specimen, and Fx and Fz data were recorded during the test. Following the test, the imaging of the entire scratch was performed automatically. The scratch was further assessed using a Bruker 3D optical microscope for the profile and dimensions of the scratch at several locations.

Figure 1. Fx and Fz plots as a function of distance (Y) for a scratch test of the nanofilm.

Results

Shown in Figure 1 are the plots of normal and lateral forces during a progressive load scratch test of the metallic nanofilm on a polymer substrate. There was a slow increase in the lateral force. The optical image of the entire scratch (top) is also shown in Figure 1. At the end, the scratch’s width was approximately 238 µm as seen from the Δx value of the image ruler at the right-top corner in Figure 1. It is shown that the film began failing at a normal load of 3.42 N, as depicted in Figure 1 with a vertical dashed line corresponding to the initiation of the semi-circular cracks on the film.

0.66 N was the corresponding Fx value and unlike the scratch test of thick film, no sharp discontinuity was there in the Fx plot at the start of the failure of the nanofilm. It is possible that the lateral force was largely dominated by the plastic deformation of the specimen at such a high normal load and not essentially due to the resistance of the nanofilm to fail. The semi-circular cracks observed at the beginning of the failure were probably formed at the trailing surface of the tungsten carbide sphere where the film was exposed to a tensile load.

Such failure could also be caused by train localization, such as necking following debonding of the nanofilm from the polymer substrate, as recommended by others.¹ Given the ductile nature of the film, no other brittle failure mode was seen under the present test conditions. When the normal load increases further, the size of the semicircular crack also increases. At approximately 6.7 N of Fz, the film exhibits a large rack that was potentially formed by a secondary crack connecting the two adjoining semicircular cracks. At such a high tensile stress, the substrate started failing too.

Figure 2. Surface profile from near start of the failure of the nanofilm showing three semicircular cracks.

Figure 3. Depth profile along the YY-line in Figure 2.

Using a Bruker 3D optical microscope (interferometer), the scratch surface was further analyzed to learn more about the dimension of the cracks thus formed. A surface profile of the semicircular cracks formed at the critical value of Fz is shown in Figure 2. Three such cracks are shown. The first crack has an estimated length of 80 µm. The depth profile along the YY-line in Figure 2 is shown in Figure 3. It is to be noted that the scratch direction was from the bottom to the top in Figure 2.

In Figure 3, the depth profile shows that there was an overall change in height of about 300 nm at the point of the scratch. The leading edge was about 100 nm below the surface, and rear edge of the first crack was nearly 200 nm above the surface. This may be caused by the spallation of the nanofilm at the rear edge before the complete failure of the film under tension, thus elevating the edge.

The leading edge, on the other hand, was pressed down since the sphere was still in contact, thus possibly causing such a discontinuity in the scratch profile. It was observed that the scratch depth below the surface was similar to the nanofilm’s thickness range. The interferometer was used to examine one of the large failed areas.

Figure 4 shows the surface profile close to the large failed area; the depth profile along the YY-line in Figure 4 is shown in Figure 5. The failed area has a 2 µm depth. Unlike the profile shown in Figure 2, Figure 5 does not show any rise in the profile at the rear edge which could be due to failure of the nanofilm owing to extreme deformation of the substrate at a higher stress level.

Figure 4. Surface profile near the end of the scratch in Figure 1, showing a large failed area.

Figure 5. Depth profile along the YY-line in Figure 4.

The scratch test was performed 10 times in order to provide statistical data on the critical value of Fz for the start of failure. Table 1 presents the values of the critical normal load for failure for all the tests. Further, the average value of the critical load was 3.39 N, with a standard deviation (SD) of 0.26 N. The presented scratch data confirmed that Bruker’s scratch test system can effectively carry out scratch testing on metallic nanofilms on a polymer substrate for flexible electronic applications.

Table 1. Critical load for the failure of metallic nanofilm in scratch.

Conclusion

Although quite difficult, scratch testing to assess the adhesion property of metallic nanofilms as the critical normal load for failure in scratch is very important for research and development and quality control of flexible electronic devices. Based on the UMT TriboLab, Bruker’s universal scratch test system can accurately evaluate such metallic nanofilms. The availability of sensors with advanced scratch tips, wide force ranges and user-friendly automated optical imaging sets this scratch test system apart from other such methods and instruments.

References

1. Li, T., Huang, Z. Y., Xi, Z. C., Lacour, S. P., Wagner, S., Suo, Z., Delocalizing strain in a thin film on a polymer substrate, Mechanics of Materials, Vol. 37 (2005) 261-273.

2. Lu, N., Suo, Z., Vlassak, J. J., The effect of film thickness on the failure strain of polymer-supported metal films, Acta Materialia, Vol. 58 (2010) 1679-1687.

3. ASTM C1624-05 (2015), Standard Test Method for Adhesion Strength and Mechanical Failure Modes of Ceramic Coatings by Quantitative Single Point Scratch Testing, ASTM International, West Conshohocken, PA, 2015

4. Bull, S. J., Failure mode maps in the thin film scratch adhesion test, Tribology International, Vol.30 (1997) 491-498.

5. Blees, M. H., Winkelman G. B., Balkenende A. R., den Toonder, J. M. J., The effect of friction on scratch adhesion testing: application to a sol–gel coating on polypropylene, Thin Solid Films, Vol. 359 (2000) 1-13.

This information has been sourced, reviewed and adapted from materials provided by Bruker Nano Surfaces.

For more information on this source, please visit Bruker Nano Surfaces.