Measurement for fracture toughness of single crystal silicon film with tensile test

doi:10.1016/j.sna.2003.10.063

Sensors and Actuators A: Physical

Volume 119, Issue 1, 28 March 2005, Pages 229-235

https://doi.org/10.1016/j.sna.2003.10.063 Get rights and content

Abstract

We developed a method of measuring the fracture toughness, which is a material constant in the macroscopic domain, of micro-machined single crystal silicon film on the (1 0 0) plane. We notched the thin film specimen on a single edge and then conducted a uniaxial tensile test to specimen failure. The average value of measured fracture toughness was 1.58 MPa m^1/2, with scatter. This is slightly higher than, but comparable to, the value for bulk silicon. Scanning electron microscope (SEM) observation of the failed specimens revealed that the fracture developed mainly along the (1 1 0) cleavage plane.

Introduction

The mechanical properties of micrometer- and nanometer-scale material are important factors to be considered when designing micro-electro-mechanical systems (MEMS), as most of these are known to exhibit dependence on the size of specimen. Single crystal silicon material is one of the most common materials used in MEMS. Considerable research has been done on evaluating Young’s modulus and the fracture strength of MEMS materials [1], [2], [3], [4]. Studies on the fracture toughness (another material property) of millimeter- to micrometer-scale silicon structures have recently been attracting a great deal of interest [5], [6], [7]. Methods used involve indentation fracture (IF) [5] and bending tests [6], [7]. These studies have provided values that are slightly higher than, but mostly comparable to bulk values. Although the tensile test is the closest representation to the Griffith crack theory in a strict sense, few known tensile methods used for measuring fracture toughness. This is primarily due to difficulties in fabrication of such small-scale test specimens, especially of the cracks.

In this paper, we describe a tensile test method for evaluating the fracture toughness of single crystal silicon film by using a micrometer-scale specimen with notch, and present the results of it.

Section snippets

Experiment

We adopted the single-edge-notched tension specimen proposed by Gross et al. [8] (shown in Fig. 1) to measure the stress intensity factor, k_I, which is a function of the applied tensile stress, σ, the notch length, a, and the geometry of the specimen, and is given by [8] $k_{I} =Yσ a$ where Y is a factor related to crack configuration and is defined as $Y=1.99−0.41λ+18.70λ^{2} −38.48λ^{3} +53.85λ^{4}$ where λ=a/W.

L, W and B (in Fig. 1) are the length, width and thickness of the specimen. Here, L 2W is indispensable to

Results

We successfully tested a total of 13 chips. The specimens were micro-fabricated on the (1 0 0) wafer, and the tensile direction was along the 〈1 1 0〉 orientation. The measured fracture stresses of the notched specimens are shown in Fig. 8. The detailed notch lengths, specimen numbers, and average fracture stresses are summarized in Table 1.

In Fig. 8, the fracture stress of notch length 0 indicates the fracture stress of non-notched film specimens, i.e., of ordinary tensile tests, which we also

SEM observation

Fig. 10, Fig. 11, Fig. 12 are SEM micrographs of typical failed film specimens. In Fig. 10, the notch length is 1.0 μm. Obviously, the crack initiated from the notch tip and advanced directly along the (1 1 0) plane, which is the maximum tensile stress plane as well as being perpendicular to the (1 0 0) film plane (Fig. 10a). The fracture section plane was flat and smooth near the notch (Fig. 10b), and followed along the 〈1 1 0〉 orientation with parallel lines until the fracture end (Fig. 10c).

This

Conclusions

We conducted tensile tests of single-edge-notched specimen on a (1 0 0) single crystal silicon film, whose tensile direction was along the 〈1 1 0〉 orientation. The results we obtained from measuring fracture toughness and the observation of fracture paths led to following conclusions.

1.
The measured average fracture toughness was 1.58 MPa m^1/2. This is slightly higher than, but comparable to, the value for bulk silicon. The average agrees with the results obtained by other researchers on micrometer- and

Xueping Li received her BS degree in 1993 and MS degree in 1996 from Nanjing University of Aeronautics and Astronautics (China). Presently, she is a PhD candidate in the Department of Micro System Engineering at Nagoya University, Japan. Her thesis research on fracture properties in single crystal silicon micro-mechanical structures is being conducted under the direction of Lecture Mitsuhiro Shikida and Prof. Kazuo Sato.

References (13)

K. Sato et al.
Tensile testing of silicon film having different crystallographic orientations carried out on a silicon chip
Sens. Actuators A
(1998)
T. Yi et al.
Microscale material testing of single crystalline silicon: process effects on surface morphology and tensile strength
Sens. Actuators A
(2000)
F. Ebrahimi et al.
Fracture anisotropy in silicon single crystal
Mater. Sci. Eng. A
(1999)
A.M. Fitzgerald et al.
Fracture toughness and crack growth phenomena of plasma-etched single crystal silicon
Sens. Actuators A
(2000)
S. Sundararajan et al.
Development of AFM-based techniques to measure mechanical properties of nanoscale structure
Sens. Actuators A
(2002)
T. Ando et al.
Tensile-mode fatigue testing of silicon films as structural material for MEMS
Sens. Actuators A
(2001)

There are more references available in the full text version of this article.

Cited by (55)

Design and fabrication of a scalable solid-propellant micro-thruster array using lab-on-PCB technology
2023, Sensors and Actuators A: Physical
The demand for microthrusters is growing with the increasing utilization of nano/microsatellite platforms in the space industry. Solid propellant microthruster arrays hold great promise for commercial applications due to their lightweight nature and optimization. However, significant challenges exist in scaling up and mass-producing traditional microthrusters, hindering their commercialization. In this paper, we propose a novel concept of a solid propellant micro thruster array and present its fabrication process using lab-on-printed-circuit-board (PCB) technology, along with feasibility studies. All constituent layers of the thruster are fabricated using a commercially available PCB, and the assembly process is conducted using standard PCB surface mount technology. The microthruster array manufactured using this technique exhibits high structural stability due to the robust physical characteristics of its components. Furthermore, the technique can be upscaled and enables cost-effective mass production of components. The ignition characteristics and propulsion performance of the microthruster array are validated through ignition and combustion experiments. A fully functional lab-on-PCB microthruster enables the fundamental integration of electronic control and propulsion systems in a satellite, offering exciting possibilities for future satellite missions.
Effect of nitrogen flow rate on the mechanical properties of (V,Mo)N thin films
2023, Surface and Coatings Technology
The objective of this research was to study the effect of nitrogen flow rate on the structure and mechanical properties of (V,Mo)N thin films. Five sets of (V,Mo)N thin films were deposited with different nitrogen flow rates ranging from 1.2 to 6.0 sccm on Si (100) substrate using unbalanced magnetron sputtering. The results showed that the N/metal ratio increased from 0.47 to 0.85 with increasing nitrogen flow rate. The microstructure of the specimen deposited at the lowest nitrogen flow rate (D12) was characterized by transmission electron microscopy, where the microstructure contained three phases, including Mo, VN and (V,Mo)N. The X-ray diffraction patterns also indicated the presence of minor Mo metal phase for the specimens deposited at nitrogen flow rate lower than 2.5 sccm. The texture of the thin films was mostly (200) and switched to random texture at nitrogen flow rate from 4.0 to 6.0 sccm. The lattice constant of the specimens was linearly related to the N/metal ratio. The specimen D12 had the highest hardness 29.3 GPa, which could be due to the existence of Mo metal phase in between (V,Mo)N phase, thereby retarding the crack propagation. By contrast, the hardness of all the other specimens showed slight variation ranging from 23.6 to 26.1 GPa, which may be attributed to the nanograin size (<15 nm) where deformation was controlled by grain boundary-mediated mechanisms. Fracture toughness (G_c) of the thin films was evaluated using internal energy-induced cracking method. The resultant G_c of the specimens, ranging from 16.4 to 30.1 J/m², linearly increased with increasing N/metal content, suggesting that the MoN bonding may be crucial on increasing fracture toughness of (V,Mo)N thin films. For specimen D12, the higher G_c may be caused by the presence of Mo metal phase that could arrest the crack propagation. The results showed that G_c of the (V,Mo)N thin films was higher than that of VN and TiN thin films, which was consistent with the theoretical predictions.
From macro- to micro-experiments: Specimen-size independent identification of plasticity and fracture properties
2021, International Journal of Mechanical Sciences
The stress-strain response of most engineering materials can be conveniently determined using well-established macro-specimens with outer dimensions of several millimeters or even centimeters. However, in the case of curved thin-walled structures or structures with spatial material property gradients, it is often impossible to extract macro-specimens. For the first group of structures the specimen dimensions are greater than the structural dimensions, while for the second group, the assumption of statistically-homogeneous material properties is not valid throughout a volume as large as the specimen gage section. A frequently encountered example is material property gradients around welds in metallic structures or castings in automotive engineering. In view of characterizing the local plasticity and fracture properties at the millimeter scale, a micro-testing technique is discussed in this study. In addition, the microscopic experimental results of 100μm thick steel alloy API X52 are compared to standard macroscopic counterparts with 2mm thickness. A new custom-made micro-testing device is employed to load uniaxial tension, notched, central hole and in-plane shear micro-specimens with a quasi-static nominal actuator strain rate of less than 1μm/s. These micro-scale experiments on a pipeline steel are performed under an optical microscope; a planar digital image correlation is used to compute the surface strain fields. The plasticity and fracture initiation model are then derived using the conventional macro- and micro-experiments. It is concluded that micro-testing provides a powerful means to characterize materials whose properties are homogeneous over a few hundred microns only.
From continuum to quantum mechanics study on the fracture of nanoscale notched brittle materials
2021, International Journal of Mechanical Sciences
Citation Excerpt :
Finally, the driven energies for fracture of small-scale and bulk single-crystal silicon obtained by current CZM simulation and previous conventional fracture mechanics [16,21-23, 56-58] are compared in Fig. 12. It should be noted that the fracture energy in previous [16,21-23, 60-62] and current studies was derived from the fracture experiment from the pre-crack [16] and the nano-notch tips [21-23, 60-62], respectively. Thus, the dimensions of concentration factors of stress fields near the crack and notch tip are completely different.
The fracture of nanoscale notched brittle materials is investigated using the multi-scale analysis of cohesive zone modeling and first-principles calculations based on the notched nano-cantilever bending experiment. The fracture of nanoscale single-crystal silicon is achieved at the nano-notch tip, and the load-deflection curve is obtained during in situ TEM experiment. On the other hand, a bilinear CZM is adopted to simulate the observed fracture from the notch tip due to the stress concentration of less than 1 nm. The CZM parameters determined from one specimen, ϕ_n=9.78 J/m², σ_max = 13.04 GPa, and $Δ_{n}^{c}$ =1.5 nm, accurately predict the fracture behaviors of all other specimens regardless of specimen sizes, indicating the robust applicability of CZM for describing the fracture of nanoscale brittle materials. In addition, first-principles calculations are performed to investigate the inherent fracture properties of single-crystal silicon from atomic and electronic viewpoints. The fracture surface energy and critical bond length for the break of atomic bonds during the fracture are compared with the cohesive energy and failure length parameter, respectively, which provides the atomistic interpretation of CZM validity for the fracture of brittle materials by the extremely small stress concentration. Finally, the comparison of cohesive energy with the fracture energy obtained by fracture mechanics of nanoscale and bulk single-crystal silicon indicates that the consumed energy is an effective linkage to quantify the fracture of brittle materials at different scales.
Gyroscopes
2020, Handbook of Silicon Based MEMS Materials and Technologies
This chapter shows how micro electro-mechanical system (MEMS) inertial sensors developed since last two decades, with focus onto gyroscopes. Inertial measurement units (IMU) in fact became since beginning of years the 2000s one of the first MEMS volume production in the consumer market. This kind of devices constantly increased their performance level and added advanced functionalities by integrating multiaxis accelerometers and gyroscopes.
Front-end silicon integration is here described reporting the new findings, adopted time by time to optimize capability and sensitivity, and next future development scenario. The key features and requirements of the gyroscope technology architecture are compared and critically analyzed.
A particular role in IMU integration is played by wafer-to-wafer bonding technique and packaging, which offers the possibility to apply to the final devices’ innovative solutions and original design inputs.
Towards diamond micro four-point probes
2019, Micro and Nano Engineering
Diamond has extreme physical properties, and it is used for critical applications in micro- and nanotechnology and nanosciences. Micro four-point probes for electrical characterization of metal and semiconductor thin-films are made from metal coated silicon microcantilevers. The limiting factors in probe lifetime are frictional wear and thermal damage from electrical overstress. Diamond has superior thermal conductivity and wear resistant properties which could improve probe lifetime. In this paper, we present wafer-level fabrication of polycrystalline diamond micro cantilevers for micro four-point probe applications. The process only requires two photolithography steps. We show great design flexibility and demonstrate extremely mechanical robust freestanding diamond cantilever up to 300 μm long. We believe that this process may find important applications beyond micro four-point probes. The next step is to implement micro four-point probes using high quality electrically conductive diamond when it becomes available.

View all citing articles on Scopus

Takashi Kasai received his BS degree in 2001 and MS degree in 2003 from the Department of Micro System Engineering at Nagoya University, Japan. Now he is an employee of OMRON Corporation, Japan.

View full text

Measurement for fracture toughness of single crystal silicon film with tensile test

Abstract

Introduction

Section snippets

Experiment

Results

SEM observation

Conclusions

Sens. Actuators A

Sens. Actuators A

Mater. Sci. Eng. A

Sens. Actuators A

Sens. Actuators A

Sens. Actuators A