Real-Time DC-dynamic Biasing Method for Switching Time Improvement in Severely Underdamped Fringing-field Electrostatic MEMS Actuators

The overall goal of the following experiment is to improve the settling time of severely under damped electrostatic fringing field micro electromechanical system actuators. This is achieved by micro fabrication of the electrostatic MAMs actuator As a second step, the substrate beneath the actuator is selectively etched, which significantly reduces squeeze film damping and enhances the severely under damped conditions. Next, dynamic biasing waveform are employed in real time in order to determine the optimum waveform parameters for fast settling time.

Results are obtained that show significant switching time improvement based on comparison to typical unit step biasing. Though the method is used to improve the switching time of electrostatic fringing field actuators, it also can be applied to fields like micro metrology to extract thin film parameters like Young's modulus, pance ratio, and biaxial tensile stress. The first step in fabricating the electrostatic fringing field MEMS beams is UV lithography.

Begin with a properly cleaned oxidized, low resistivity silicon substrate, 25 millimeters by 50 millimeters and 0.5 millimeters thick. In this video, schematic depictions of the substrate cross-section and top view will show the progress of the fabrication. Place the substrate on the chuck of a photo resist spinner spin coat, hexa methyl dine, and then positive photo.

Resist each at 3, 500 RPM for 30 seconds. Transfer the sample to a hot plate set at 105 degrees Celsius and soft bake the photo resist for 90 seconds. This is the schematic view of the sample at this point.

Once this is done, move the sample to a mask aligner. Employ the device, mask a simplified version of which is shown, and expose the sample to UV radiation with a wavelength of 350 to 450 nanometers and an exposure energy of 391 millijoules per centimeter squared. Next, take the slide to two prepared beakers, one with tetraethyl ammonium hydroxide based developer, and another with deionized water each with sufficient volume to submerge the sample, submerge the exposed sample in the developer for 12 to 20 seconds and gently agitate it.

Then quickly remove the sample and submerge it in the rinse water for 10 seconds. After the sample is thoroughly rinsed and then dried with nitrogen, inspect the sample under a microscope to see if further development is necessary. This figure represents the sample when development is completed.

Proceed with wet etching with buffered hydrofluoric acid to remove the oxide layer. Then remove the photoresist using acetone and isopropyl alcohol after removal of the photoresist. The next step is a chemical etch with tetraethyl ammonium hydroxide.

Use a hot plate with a thermal couple to maintain temperature, a magnetic stir bar, a clean four liter peaker, and a Teflon basket that can be supported in the beaker. Pour tetraethyl ammonium 25%by weight up to the two liter mark of the beaker, and add the stir bar affix thermocouple in the solution. Then preheat the solution to 80 degrees Celsius once the solution is at 80 degrees Celsius.

Place the sample in the basket, hang the basket in the solution from the lip of the beaker. Be sure to leave room for the magnetic stir bar to rotate. Set the rotation rate of the stir bar to 400 RPM and etch to a depth of four to five micrometers.

After about 15 minutes, properly rinse and dry the sample and check the step height with a profilometer. This figure depicts the sample when the step height has been reached. The next step is to remove the silicon dioxide.

Have ready a Teflon beaker with hydrofluoric acid, 49%by volume sufficient to cover the sample and a similarly prepared Teflon beaker with deionized water. Place the sample in the hydrofluoric acid until all of the silicon dioxide is removed less than 30 seconds. Transfer the sample to the deionized water and leave it there for 10 to 20 seconds.

Continue with further rinsing and also remove organic residues from the sample. Take the clean and dry sample and perform wet thermal oxidation to grow 500 nanometers of silicon dioxide on its surface. Here is the representation of the sample after the growth step when the silicon dioxide layer is completed, once again, spin code hexa methyl xylazine followed by positive photoresist onto the sample.

After soft baking the photo resist, use a mask aligner to expose the sample to 350 to 400 nanometer UV radiation. With an exposure energy of 391 millijoules per centimeter squared, the mask depicted is simplified. Next, developed a sample in a tetraethyl ammonium hydroxide based developer.

Complete the etching with buffered hydrofluoric acid and remove the photo. Resist The sample is now ready for sputter deposition of the electro plating seed layer sputtered deposit. 20 nanometers of titanium followed by 100 nanometers of gold.

This depicts the sample after it has been retrieved. Now spin coated again with hexa methyl DIY at 3, 500 RPM for 30 seconds. Then positive photo resist at 2000 RPM for 30 seconds on a mask aligner align and expose the sample to 350 to 450 nanometer uv.

With an exposure energy of 483 millijoules per centimeter squared, developed a sample in a tetraethyl ammonium hydroxide based developer. After rinsing it thoroughly, inspect the sample with a microscope to determine if further development is necessary. Once developing is complete, electroplate the gold mems beam by first filling a clean one liter glass beaker with 700 milliliters of gold electroplating solution.

Place the beaker on a hot plate and set the hot plate to 60 degrees Celsius. Use a thermocouple to ensure the temperature stays at the desired temperature. When the temperature is 60 degrees Celsius, attach the sample to a fixture that also holds thermocouple and a stainless steel anode.

Electroplate the sample using a current density of two milliamps per centimeter squared until the necessary time has elapsed. About two minutes, turn off the current supply and remove the electrodes. To retrieve the sample.

Verify the completion of electroplating. Using a profilometer and microscope etch the photo. Resist mold by first filling a glass speaker with a dedicated photo resist stripper.

Place this on a hot plate and warm it to 110 degrees Celsius. When the temperature is reached, submerge the sample in the solution for one hour. At the end of the hour, remove the beaker from the hot plate and allow the solution and sample to cool to room temperature.

These sketches represent what should be seen after the sample is cleaned. The yellow region on the right represents the deposited gold layer. The gold regions represent the electroplated beams.

After roughening the surface, immerse the sample in a Teflon beaker with gold etching for 30 to 45 seconds On completion, quickly terminate the etch by submersing the sample in deionized water for 10 to 20 seconds when all the gold is removed, and after more roughening, submerge the sample in buffered hydrofluoric acid etch at room temperature for three to six seconds, rinse dry and inspect the etch to make sure all the titanium is removed from the exposed areas as shown here schematically. As a final step, perform a dry isotropic xenon di fluoride edge. This will selectively remove the silicon and release the gold fixed, fixed beams.

This represents the final beam configuration after fabrication of the fixed beams. The next step is experimental validation. Place the sample on the chuck of a DC probe station and secure it.

Next, use the microscope of the probe station to precisely position the tungsten probe tips. Place the live signal probe tip over the pad for the fixed fixed beam. Place the ground probe tip over the pad for the pull down electrodes over the DC biasing pads of the device.

Next, set the parameters of the waveform on the function generator. Based on calculation, pass the output of the function generator to a high speed, high voltage linear amplifier. Monitor the output with the digital oscilloscope with a sampling rate of 300 megahertz.

With the function generator off, connect the output of the linear amplifier to the DC manipulators. Now position the laser doppler barometer over the sample and turn on the laser. Use the integrated microscope to identify the desired beam and focus the laser on its center.

Select the sampling rate output and measurement modes With the preparations complete. Apply the biasing signal to the MEMS bridges. Start tuning the timing and voltage parameters on the function generator Work to achieve minimal beam oscillation on the pull down and release operations.

Once the optimal values are found, turn off the biasing signal and the continuous laser doppler barometer measurement mode. Lift the probe tips from the biasing pads. Then run the device in single scan mode using a viter as described in the accompanying text protocol.

Return the DC probe tips to the biasing pads of the MEMS bridge. Activate the signal scan and capture the displacement versus time data. Here are the beam deflections is a function of time for a MEMS bridge in response to a 60 volt input bias for four different conditions.

Step release responses are shown in black and exhibit oscillation. The dynamic release measurements shown in red are made after timing and voltage measurements are tuned. These show the oscillations are largely removed.

Once found, the dynamic waveform is useful for all gaps, not just the one associated with a 60 volt input bias. Here, the pull down response is shown for several gap heights corresponding to different input bias voltages. In each case, oscillations are limited.

Similarly, here are the release responses in every case. Note that virtually all oscillations are removed. Please remember that working with chemicals like hydrofluoric acid and TMAH can be extremely hazardous.

Therefore, taking precautions like wearing personal protective equipment should be taken while performing this procedure.