Employee Locator | Index | Site Map
Home > Mission > ST&E > Microsystems > MEMS > Capabilities

Microsystems Home


SANDIA MEMS

Capabilities



MolTun (Molded Tungsten)™

Overview

The oxide molded tungsten process can be used for the fabrication of complex multilevel structures which can be relatively thick. The steps in the process are the deposition of a sacrificial oxide, patterning etching of the oxide, filling of the resulting mold by a blanket film of tungsten using chemical vapor deposition, and then removal of excess tungsten and planarization through chemical mechanical polishing. We have termed this process MOLTUN™ which is short for Molded Tungsten. This process has been used to demonstrate complex multilevel devices such as micro-mass-analysis systems up to 25 microns thick and novel latching relays which take advantage of the inherent tensile stress of the tungsten. Benefits of this process include high Z stiffness, high force actuation, high capacitance, maintenance of planarity and reduction or compensation for stress. Possible applications for MolTun™ include actuators, miniature mass analyzer, adaptive optics, and inertial sensors. Using current microfabrication process and materials it is possible to fabricate a wide very large variety of useful devices.

Micro-Mass-Analysis Systems Applications

MolTun™ has been used in Micro-Mass-Analysis Systems resulting in large (up to 106) arrays of cylindrical ion traps consisting of 4 separate electrode layers each separated by gaps. The devices are thicker than those achievable using traditional surface micro-machining and are far more complex than those achievable using deep trench etching. The pictures below show the highly complex W parts. The thickest parts fabricated for this project had ~14 layers of tungsten, 8 of which went into the fabrication of the ring electrode. The total thickness of the structures was ~25 microns. This is considerably thicker than typical polysilicon micromachined structures.

   
SEM micrographs of partially completed micro-mass-analysis systems.

Micro-Latching Relays

MolTun™ has also been used in micro-latching relays in which the design takes advantage of the internal tensile stress of the released W. The micro-mass analysis system is mechanically a static device and the residual tensile stress of the W is dealt with by simply anchoring the array firmly to the substrate. In the micro-latching relay this tensile stress is critical to the functioning of this mechanical device. The general design is based on having two pairs of opposing thermal actuators. The first picture below shows a typical released chevron type thermal actuator fabricated in the MOLTUN™ process. Initially the arms of the thermal actuator were at 4 degrees to the line between the pad centers. Upon release, the residual tensile stress of the W has pulled the arms of the thermal actuator straight. Our results show that the displacement can be considerable, tens of microns depending on geometry, and that angles greater than 4 degrees would be required to completely "relax" the tensile stress associated with these released W devices. In our latching relay design we place two orthogonal pairs of identical thermal actuators in opposition to each other, as seen in the second picture below. Since the system is symmetrical the overall displacement of the central shuttle is zero at rest. Passing a current to heat one of the heater arms causes it to expand and allows the opposing thermal actuator to reduce its internal stress by pulling the central shuttle towards it. The forces exerted in this design are always "pulling" in nature. In a conventional thermal actuator design the force generated is typically "pushing" in nature and when the force becomes sufficiently large the thermal actuator buckles and it becomes impossible to increase the actuation force. In our "pulling" W design it is impossible for the system to relax the force exerted by deforming into a buckling mode.


Top view optical micrograph of a released chevron thermal actuator fabricated using the MOLTUN™ process. Initially the arms of the thermal actuator were at 4 degrees to the line between the center of the two anchor pads. The residual stress in the W resulting from the thermal expansion mismatch between the Si and the W is sufficient to pull the arms straight.


Oblique angle SEM micrograph showing the latching relay design developed for the MOLTUN™ process. In this design two pairs of identical thermal actuators are arranged opposing each other and attached to each other through a suspension and a suspended plate. Since the system is symmetrical the overall displacement of the central shuttle is zero. Actuating one of the chevron thermal actuators allows the opposing thermal actuator to relax and pull back. Having two sets of orthogonal pairs of actuators enables X-Y translation of the central shuttle.


The opposing thermal actuation mechanism translates the top portion of the shuttle so that the two teeth engage. Once the teeth are engaged the actuation mechanism can be powered down and the structure remains latched in position. Two such structures electrically connect two pads through the moving shuttle.

Contact MEMS at Sandia: memsinfo@sandia.gov



MEMS Image Gallery



MEMS Bibliography
MEMS Bibliography



Design Manual …
DRC Error Descriptions
Design Review Checklist



MEMS Bibliography
MEMS Brochure