Olivier N. Pierron

Fatigue rates of monocrystalline silicon thin films in harsh environments: Influence of stress amplitude, relative humidity, and temperature

This study investigates the separate influence of stress, temperature, and relative humidity (RH) on the fatigue behavior of 10-μm-thick, monocrystalline silicon (Si) films at 40 kHz, under fully reversed loading. The fatigue rates are most sensitive to stress, with four orders of magnitude decrease from 3.2 to 1.5–2 GPa, confirming a size effect associated with the fatigue behavior of Si under bending load. The fatigue rates are also much more sensitive to RH than temperature or partial pressure of water, indicating that the effective environmental parameter is the adsorbed water layer. The implications on the relevant fatigue process(es) are discussed.

A versatile microelectromechanical system for nanomechanical testing

This letter presents a microelectromechanical system (MEMS) material testing setup that relies on electronic measurements of nanospecimen elongation. Compared to previously demonstrated MEMS that rely on high magnification images to measure elongation, this MEMS is more versatile, allowing both in situ and ex situ testing of nanomaterials with high accuracy and precision. We describe and characterize the MEMS device and illustrate its mode of operation with a successful ex situ uniaxial tensile test of a nanocrystalline nickel nanobeam. The combination of ex situ and in situ nanomechanical tests will enable a thorough investigation of critical properties pertaining to the reliability of nanosystems.

Comparison of the Stress Distribution and Fatigue Behavior of 10- and 25-

The stress distribution and fatigue behavior of nominally identical kilohertz fatigue resonators with two different thicknesses, 10 and 25 μm, was compared in this study. The results highlight the non-uniform 3-D stress distribution of the micron-scale notched cantilever beams that depends on the thickness. The areas corresponding to the first principal stress being within 2% of the maximum value are much smaller than the overall notch area and are a function of device thickness. It is also shown that the non-negligible influence of small, nanometer-scale geometrical variations in the dimensions of nominally identical devices on the maximum stress values can be accounted for by measuring the devices resonant frequency (f0). The observed scatter in the fatigue results of these microresonators is in part associated with the challenge in accurately calculating the local stress amplitudes. Despite that large scatter, the fatigue behavior of the 10 and 25 μm thick devices is similar. Particularly, the overall relative decrease rates in f0 are well related to fatigue life (Nf) and can be used to predict Nf within a factor of 5, for Nf ranging from 104 to 1010 cycles.

\mu\hbox{m}

This paper presents a testing methodology for measuring the low cycle fatigue properties of single-crystal silicon thin films using kHz-frequency resonators. The dynamic behavior of the fatigue structures is thoroughly characterized to allow accurate measurements of stresses (±0.1 GPa) and fatigue lives (±250 cycles). The tests consist of applying successive bursts of small numbers of cycles (as low as ~500 cycles) and measuring the resonant frequency in between each burst. Continuous damage accumulation, beginning after the first burst, is observed based on the decrease in resonant frequency of the resonant structure.

-Thick Deep-Reactive-Ion-Etched Si Kilohertz Resonators

A singular critical onset strain value has been used to characterize the strain limits of barrier films used in flexible electronics. However, such metrics do not account for time-dependent or environmentally assisted cracking, which can be critical in determining the overall reliability of these thin-film coatings. In this work, the time-dependent channel crack growth behavior of silicon nitride barrier films on poly(ethylene terephthalate) (PET) substrates is investigated in dry and humid environments by tensile tests with in situ optical microscopy and numerical models. The results reveal the occurrence of environmentally assisted crack growth at strains well below the critical onset crack strain and in the absence of polymer-relaxation-assisted, time-dependent crack growth. The crack growth rates in laboratory air are about 1 order of magnitude larger than those tested in dry environments (dry air or dry nitrogen). In laboratory air, crack growth rates increase from ∼200 nm/s to 60 μm/s for applied stress intensity factors, K, ranging from 1.0 to 1.4 MPa·m1/2, below the measured fracture toughness Kc of 1.8 MPa·m1/2. The crack growth rates in dry environments were also strongly dependent on the prior storage of the specimens, with larger rates for specimens exposed to laboratory air (and therefore moisture) prior to testing compared to specimens stored in a dry environment. This behavior is attributed to moisture-assisted cracking, with a measured power law exponent of ∼22 in laboratory air. This study also reveals that much larger densities of channel cracks develop in the humid environment, suggesting an easier initiation of channel cracks in the presence of water vapor. The results obtained in this work are critical to address the time-dependent and environmental reliability issues of thin brittle barriers on PET substrates for flexible electronics applications.

Low-cycle fatigue testing of silicon resonators

Abstract The fatigue properties of ultrathin protective coatings on silicon thin films were investigated. The cohesive and delamination fatigue properties of 22 nm-thick atomic-layered-deposited (ALD) titania were characterized and compared to that of 25 nm-thick alumina. Both coatings were deposited at 200 °C. The fatigue rates are comparable at 30 °C, 50% relative humidity (RH) while they are one order of magnitude larger for alumina compared to titania at 80 °C, 90% RH. The improved fatigue performance is believed to be related to the improved stability of the ALD titania coating with water compared to ALD alumina, which may in part be related to the fact that ALD titania is crystalline, while ALD alumina is amorphous. Static fatigue crack nucleation and propagation was not observed. The underlying fatigue mechanism is different from previously documented mechanisms, such as stress corrosion cracking, and appears to result from the presence of compressive stresses and a rough coating–substrate interface.

Environmentally Assisted Cracking in Silicon Nitride Barrier Films on Poly(ethylene terephthalate) Substrates

A microresonator-based interfacial fatigue testing technique was used to investigate the subcritical delamination of atomic-layer-deposited alumina coatings along the sidewalls of deep-reactive-ion-etched monocrystalline silicon thin films. This technique ensures loading conditions relevant to microelectromechanical system devices, including kHz testing frequency and fully reversed cyclic stresses. Four different coating thicknesses (4.2, 12.6, 25, and 50 nm) were investigated in two environments (30 °C, 50% relative humidity (RH) and 80 °C, 90% RH). Fatigue damage, in the form of channel cracks and delamination of the alumina coating, was found to accumulate slowly over more than 1 × 10(8) cycles. The average delamination rates increase with increasing energy release rate amplitude for delamination, modeled with a power law relationship. In the harsher environment, the rates are roughly 1 order of magnitude higher. Additionally, a few tests under static load were conducted for which no delamination (or crack growth) occurred, demonstrating that the governing interfacial fatigue mechanism is cycle-dependent.

Comparison of the cohesive and delamination fatigue properties of atomic-layer-deposited alumina and titania ultrathin protective coatings deposited at 200 °C

We introduce an external-load-assisted thin film channel crack growth technique to measure the subcritical crack growth properties of thin films (i.e., crack velocity, v, versus the strain energy release rate, G), and demonstrate it using 250-nm-thick SiNx films on poly(ethylene terephthalate) substrates. The main particularity of this technique is that it requires a polymer substrate to allow loading to large strains (in order to induce channel cracking) without substrate fracture. Its main advantages are to provide a full v-G curve with a single specimen while relying on a simple specimen preparation and straightforward crack growth characterization. Importantly, the technique can be employed for a much larger range of thin films compared to the residual-stress-driven, thin film channel crack growth tests, including ultrathin films and thin film with residual compressive stresses. The restrictions to a proper use of this technique, related to the (visco)plastic deformation of the substrate, are discussed.

Interfacial cyclic fatigue of atomic-layer-deposited alumina coatings on silicon thin films.

Signature parameters, such as true activation volume and effective stress, are often characterized to identify the governing plastic deformation mechanisms, including that of nanocrystalline metals. The accurate measurement of these parameters using transient tests was recently questioned for nanocrystalline metals, in which grain-boundary-based mechanisms can concurrently occur with dislocation glide. Here, we demonstrate the use of a microelectromechanical systems (MEMS) device to measure true activation volume and effective stress based on repeated stress relaxation and stress dip experiments, respectively. The technique was demonstrated on 100-nm-thick nanocrystalline Au microbeams. These miniaturized tests open up the possibility of observing the mechanisms directly under a transmission electron microscope, and providing a direct link between these measured parameters and the governing mechanisms. [2016-0306]

Note: A single specimen channel crack growth technique applied to brittle thin films on polymer substrates

This Letter presents a quantitative in situ scanning electron microscope (SEM) nanoscale high and very high cycle fatigue (HCF/VHCF) investigation of Ni microbeams under bending, using a MEMS microresonator as an integrated testing machine. The novel technique highlights ultraslow fatigue crack growth (average values down to ∼10-14 m/cycle) that has heretofore not been reported and that indicates a discontinuous process; it also reveals strong environmental effects on fatigue lives that are 3 orders of magnitude longer in a vacuum than in air. This ultraslow fatigue regime does not follow the well documented fatigue mechanisms that rely on the common crack tip stress intensification, mediated by dislocation emission and associated with much larger crack growth rates. Instead, our study reveals fatigue nucleation and propagation mechanisms that mainly result from room temperature void formation based on vacancy condensation processes that are strongly affected by oxygen. This study therefore shows significant size effects governing the bending high/very high cycle fatigue behavior of metals at the micro- and nanoscales, whereby the stress concentration effect at the tip of a growing small fatigue crack is assumed to be greatly reduced by the effect of the bending-induced extreme stress gradients, which prevents any significant cyclic crack tip opening displacement. In this scenario, ultraslow processes relying on vacancy formation at the subsurface or in the vicinity of a crack tip and subsequent condensation into voids become the dominant fatigue mechanisms.