Measurement of Energy Loss in Thin Films Using Microbeam Deflection Method Part 1

ABSTRACT

A technique developed for studying the energy loss behavior of submicron to nanometer scale thin metal films on substrate is presented. The test microstructure was designed the triangular cantilever beam and fabricated by the standard CMOS processes, which can improve stress distribution non-uniform problem and the thickness regime of deposited metal thin film on its surface could reduce to several nanometers. In order to reduce the measure error and calculation complex due to the contact force, the driving system was used electrostatic force to making the paddle cantilever beam bend and the deflection of paddle cantilever beam due to the electrostatic force was measured by a capacitance change. The deflection of the paddle beam can be measured from the capacitance value. A force equilibrium calculate method (include sample compliance force, force due to the film, force due to the gravity and electrostatic force) could determine the stress and strain of the deposited films easily. The anelastic behavior and internal friction of 200~500 nm Al thin film were studied using the dynamic frequency response of the paddle structure generated by electrostatic force under vacuum pressure. The result show the measurement system used here can accurately measures the loss mechanism of thin film using dynamic response which give potential to study the grain boundary motion and dislocation motion in the nano-scale thin films.

Introduction

With the development of the micro-electro-mechanical systems (MEMS) technology, the system and device design required further miniaturization in order to increase the performance need and cost efficiency. As a result, the mechanical properties of sub micron and nano scale thin films have become one of the most important issues. In MEMS applications, the static mechanical properties such as residual stress, modulus and fractural toughness are keys for the design protocol. Moreover, dynamic properties of metal thin films as a function of the vacuum pressure can be pivotal. However, due to the difficulties on measurement techniques, a simple and accurate measurement arrangement cannot be fulfill [1].


Many methods used to measure the mechanical properties of the thin film have been proposed in previous studies. The results obtained from different measurement techniques were vary widely for nominally identical samples due to the difficulty with the techniques [2, 3]. Example of the traditional micro-beam bending test used nano-indentation measure the relation between the applied force and the deflection, but the indenter tip touching directly the sample surface may break the thin film. Each method has the difficult techniques on itself need to overcome.

In the literature reviews on dynamic damping responds of materials, many anelastic mechanisms had investigated in bulk material [3], but rarely in thin films. In order to understanding the accurate static and dynamic response of the thin film materials, the energy dissipation study through simply damping response of thin film on substrate was performed. Here we developed a method that can be used to measure the energy loss of thin metal films with thickness less than 100 nm. The test specimen was designed to deposit on a novel triangle shape "paddle" beam in order to provide uniform plane strain distribution. When the sample reached the desired thickness, the tested thin film on the top surface can then be tested for measuring its static and dynamic mechanical properties.

Experimental

Previously, a paddle cantilever beam was proposed [4, 5]. This paddle like beam is different from the traditional parallel-sided cantilever beam; it was designed to have triangular side shape to provide a uniform stress distribution. The connected square plate is used as an electrode to induce the electrostatic force to make the deflection during the tests. The schematic and dimensions of paddle sample is shown in figure 1.

Schematic and dimensions of paddle sample

Figure 1- Schematic and dimensions of paddle sample

The sample fabrication procedure was using standard clean room processing and the sample frame dimensions are 20mm square, the length of triangular beam from the fixed end to the free end connected to the paddle plate is 3mm. The area of the paddle plate is 25mm2. The thickness of constant stress beam is 40^m after complete fabricate processing. Each of the paddle samples were fabricated through standard semiconductor fabricate processing. A four inches double sides polished silicon wafer using the RCA clean process for removing particles and organisms due to environment then grown silicon nitride about 200nm on both wafer surfaces Using low pressure chemical vapor deposition (LPCVD). Photoresist layer on both sides of wafer were patterned using two aligned mask. Anisotropic etching (ICP-RIE) was used to etching silicon nitride and to remove the photoresist. The un-protect regime on the silicon wafer was etched in 30%wt. KOH solution at 85°C until the beam thickness reaching 40^m. Finally, remove the etching barrier layer. The metal film, serve as conductive layer, was then deposit on the bottom surface of the silicon. The metal film that interested to be measured was then deposit on the top surface of the sample. The complete fabrication process flow is shown in figure 2.

The fabrication flow of the paddle sample

Figure 2- The fabrication flow of the paddle sample

The experiment is to deflect "paddle" cantilever beam sample using the electrostatic force at bottom and then using capacitance measurement on the other side of the deflected plate to measure its deflection with respect to the force and so on. This technique has allowed testing of thin films at the desired length scales with the thickness as few hundred nanometers to less than 10 nanometers and also maintained consistent preparation and experimental procedures.

The electrical driving circuits system was based on the charging parallel-plate principle when the two parallel planes shift produces the electrical field. Here, the two charging plates in the system were both the bottom surface of the sample and the electrode underneath the paddle plate. It is defined the electrostatic force

tmp10-122_thumb

where, Fe is the electrostatic force, e0 is the dielectric constant in the vacuum, V is the applied voltage, d is the distance between the two parallel plates and A is the effective area of the parallel plates.

When the paddle beam is bending, the distance between the two plates is not the constant but is a function of x. Therefore, equation (1) is rewritten as equation (2).

tmp10-123_thumb

where Fe is the electrostatic force, e0 is the dielectric constant in the vacuum, lb is the length of the paddle cantilever beam, lp is the length of the paddle plate, V is the applied voltage, de is the distance between the under surface of the sample and the electrode, and yb is the position of the paddle cantilever end. The symbol in the cantilever beam and the test sample structure are shown in the figure 3.

The symbol definition of optical measure system. Integrating equation (2) and rewriting it as equation (3), the electrostatic force is defined as

Figure 3- The symbol definition of optical measure system. Integrating equation (2) and rewriting it as equation (3), the electrostatic force is defined as

tmp10-125_thumb

During the experiment, we apply an electrostatic force to the specimen and measure the capacitance change. The deflection of the paddle beam can be measured from the capacitance value. The sample chip is mounted together with a guard-ringed capacitor electrode as shown in Figure 4. A spacing of 25 to 125um to the window frame chip surface around the paddle structure is defined by a metallic spacer.

The electrostatic deflection "paddle"

Figure 4- The electrostatic deflection "paddle"

A second electrode is mounted below the paddle plate. This electrode is used for electrostatic deflection of paddle. The whole chip is at DC ground but driven at 100 kHz with amplitude of a few volts. That provides a displacement current to central electrode of the capacitor plate which is proportional to the capacitance, and hence inversely proportional to the gap. Depending on the spacing selected, the capacitance is between 2 and 4 pF. The measurement of the capacitance can be made to a precision of approximately 0.1 fF so that paddle spacing changes of 50 nm are readily determined. The paddle can be pulled up with a DC voltage on the guard-ringed electrode or pulled down with a DC voltage on the lower electrode. The capacitance measurement can be made with a time resolution of ± 10 msec. For the electronic setup of the capacity measurement, a sine-wave generator at 100 kHz is applied to the film simultaneously measuring capacity of the paddle capacitor while a second generator drives at the same frequency for a test capacitor which has a known capacity. The two units are coupled (one master, one slave) and have a phase shift of 1800. Figure 5 show the circuits. The 1800 out of phase currents from the two generator-capacity pairs are summed at input of change sensitive preamplifier. The amplified sum is measured with lock-in amplifier with the reference signal from one of the frequency generators.

Electronic setup for the capacity measurement

Figure 5- Electronic setup for the capacity measurement

The measurement is controlled by PC through National Instrument LabVIEW program. The control electronics include a controller, amplifier and waveform generator. Monitored signals are conditioned and then fed into an A/D board which is located in a PC. Data acquisition is performed with LabVIEW software. During sample testing, it is placed inside the vacuum chamber. After the sample is being locked inside the system, then wait until the system is reaching the thermal equilibrium and the capacitor read out is clear, the test can be perform.

Next post:

Previous post: