WO2000071981A1

WO2000071981A1 - Micromachined displacement sensors and actuators

Info

Publication number: WO2000071981A1
Application number: PCT/IL2000/000268
Authority: WO
Inventors: Dan Haronian
Original assignee: Ramot University Authority For Applied Research & Industrial Development Ltd.
Priority date: 1999-05-19
Filing date: 2000-05-10
Publication date: 2000-11-30
Also published as: AU4427800A; IL130045A0

Abstract

A micromachined displacement sensor chip (100) including a reference frame (102); at least one suspended waveguide element (104) having an in-plane degree of freedom being integrally formed with the reference frame (102); a light source being integrally formed with the reference frame and being optically coupled to the at least one suspended waveguide element at one end thereof (106); and a light sensor being integrally formed with the reference frame and being optically coupled to the at least one suspended waveguide element at another end thereof; such that when the at least one suspended waveguide element is subjected to an external force, an in-plane displacement of the at least one suspended waveguide element is monitorable by the light sensor due to light modulation.

Description

MICROMACHINED DISPLACEMENT SENSORS AND ACTUATORS

FIELD AND BACKGROUND OF THE INVENTION

The present invention relates to micromachined displacement sensors and actuators and, more particularly, to such sensors and actuators employing in-plane (x-y) degree of freedom combined with on-chip detection of displacement.

Silicon micromachined inertial sensors:

The following presents a brief review of silicon micromachined accelerometers and gyroscopes which is base mainly on a review by Navid Yazdi, Farrokh Ayazi, and Khalil Najafϊ, entitled "Micromachined Inertial Sensors", Proceedings of the IEEE, Vol. 6, No. 8, August 1998, which is incoφorated herein by reference.

Inertial sensors have seen a steady improvement in their performance, and today, microaccelerometers can resolve accelerations in the micro-g range, while the performance of gyroscopes has improved by a factor of ten every two years during the past eight years. This impressive drive to higher performance, lower cost, greater functionality, higher levels of integration and higher volume will continue as new fabrication, circuit, and packaging techniques are developed to meet the ever increasing demand for inertial sensors.

Micromachined inertial sensors, consisting of accelerometer and gyroscopes, are one of the most important types of silicon-based sensors. Microaccelerometers alone have the second largest sales volume after pressure sensors, and it is believed that gyroscopes will soon be mass produced at similar volumes. The large volume demand for accelerometers is due to their automotive applications, where they are used to activate safety systems, including air bags, to implement vehicle stability systems and electronic suspension. However, the application of accelerometers covers a much broader spectrum where their small size and low cost have even a larger impact. They are used in biomedical applications for activity monitoring; in numerous consumer applications, such as active stabilization of picture in camcorders, head-mounted displays and virtual reality, three- dimensional mouse, and sport equipment; in industrial applications such as robotics and machine and vibration monitoring; in many other applications, such as tracking and monitoring mechanical shock and vibration during transportation and handling of a variety of equipment and goods; and in several military applications, including impact and void detection and safing and arming in missiles and other ordinance. High-sensitivity accelerometers are crucial components in self-contained navigation and guidance systems, seismometry for oil exploration and earthquake prediction, and microgravity measurements and platform stabilization in space. The impact of low-cost, small, high-performance, micromachined accelerometers in these applications is not just limited to reducing overall size, cost, and weight. It opens up new market opportunities such as personal navigators for consumer applications, or it enhances the overall accuracy and performance of the systems by making formation of large arrays of devices feasible. Micromachined gyroscopes for measuring rate or angle of rotation have also attracted have also attracted a lot of attention during the past few years for several applications. They can be used either as a low-cost miniature companion with micromachined accelerometers to provide heading information for inertial navigation purposes or in other areas, including automotive applications for ride stabilization and rollover detection; some consumer electronic applications, such as video-camera stabilization, virtual reality, and inertial mouse for computers; robotics applications; and a wide range of military applications. Conventional rotating wheel as well as precision fiber-optic and ring laser gyroscopes are too expensive and too large for use in most emerging applications. Micromachining can shrink the sensor size by orders of magnitude, reduce the fabrication cost significantly, and allow the electronics to be integrated on the same silicon chip.

Micromachined accelerometers: An accelerometer generally consists of a proof mass suspended by compliant beams anchored to a fixed frame. The proof mass has a mass of M, the suspension beams have an effective spring constant of K, and there is a damping factor (D) affecting the dynamic movement of the mass. The accelerometer can be modeled by a second-order mass-damper-spring system. External acceleration displaces the support frame relative to the proof mass, which in turn changes the internal stress in the suspension spring. Both this relative displacement and the suspension-beam stress can be used as a measure of the external acceleration.

By using Newton's second law and the accelerometer model, the mechanical transfer function can be obtained: x(s)

H(s) = a(s) ^■> D K , ω_r , ^; s + — s + — S- + — S + CO

M M Q _{( ] )} where a is the external acceleration, x is the proof mass displacement, ω_r = (KJM)-2 is the natural resonance frequency, and Q = (KM)^~2/D is the quality factor. The static sensitivity of the accelerometer is shown to be:

^Xs_la c ₌ M ₌ 1 a K ω) ₍₂₎

As evident, the resonance frequency of the structure can be increased by the spring constant and decreasing the proof mass, while the quality factor of the device can be increased by reducing damping and by increasing proof mass and spring constant. Last the static response of the device can be improved by reducing its resonant frequency-

The primarily mechanical noise source for the device is due to Brownian motion of the gas molecules surrounding the proof mass and the Brownian motion of the proof-mass suspension or anchors. The total noise equivalent acceleration (TNEA) [m/s(Hz)^~2) is:

where Kβ is the Boltzmann constant and T is the temperature in Kelvin. Equation 3 clearly shows that to reduce mechanical noise, the quality factor and proof mass have to be increased.

In the general case, the proof-mass motion can have six degrees of freedom. But typically in a unidirectional accelerometer, the geometrical design of the suspension is such that one of these is dominant and the device has low off-axis sensitivity. The cantilever support has been one of the early popular suspension support designs due to its simplicity, lower spring constant, and internal stress relief of the beams. However, this configuration results in a larger off-axis sensitivity unless the device is fully symmetric. Also, symmetric fullbridge supports result in a very low off- axis sensitivity, and by using a crab-leg or folded-beam configuration in a full-bridge support, the residual stress of the beams can also be relieved. The general design of accelerometers can be performed using the above equations, as well as mechanical relations describing the spring constant and damping factor as a function of device geometry and ambient pressure. Further, the device first-order design optimization can be obtained using the same equations, while the final accelerometer design can be simulated and optimized using commercially available finite element method or dedicated microelectromechanical systems (MEMS) software packages. Accelerometers are typically specified by their sensitivity, maximum operation range, frequency response, resolution, full-scale nonlinearity, offset, off-axis sensitivity, and shock survival. Since micromachined accelerometers are used in a wide range of applications, their required specifications are also application dependent and cover a rather broad spectrum. For instance, for microgravity measurements devices with a range of operation greater than ±0.1 g, a resolution of less than 1 μg in a frequency range of zero frequency to 1 Hz are desired, while in ballistic and impact sensing applications, a range of over 10.000 g with a resolution of less than 1 g in a 50 kHz bandwidth is required. A variety of transduction mechanisms have been used in microaccelerometers. Some of the more relevant and useful approaches will be reviewed here.

Piezoresistive accelerometers: The first micromachined, and one of the first commercialized, microaccelerometers were piezoresistive- These accelerometers incoφorate silicon piezoresistors in their suspension beam. As the support frame moves relative to the proof mass, the suspension beams will elongate or shorten, which changes their stress profile and hence the resistivity of their embedded piezoresistors. These piezoresistors are generally placed at the edge of the support rim and proof mass, where the stress variation is maximum. Therefore, a resistive half- bridge or full bridge can be formed by employing two or four piezoresistors. The main advantage of piezoresistive accelerometers is the simplicity of their structure and fabrication process, as well as their readout circuitry, since the resistive bridge generates a low output-impedance voltage. However, piezoresistive accelerometers have larger temperature sensitivity, and smaller overall sensitivity compared to capacitive devices, and hence a larger proof mass is preferred for them.

Capacitive accelerometers: In the presence of external acceleration, the support frame of an accelerometer moves from its rest position, thus changing the capacitance between the proof mass and a fixed conductive electrode separated from it with a narrow gap. This capacitance can be measured using electronic circuitry. Silicon capacitive accelerometers have several advantages that make them very attractive for numerous applications ranging form low-cost, large volume automotive accelerometer to high- precision inertial-grade microgravity devices. They have high sensitivity, good DC response and noise performance, low drift, low temperature sensitivity, low-power dissipation, and a simple structure. However, capacitive accelerometer can be susceptible to electromagnetic interference (EMI), as their sense node has high impedance.

Some of the most widely used structures for capacitive accelerometers are vertical and lateral structures. Many capacitive accelerometers utilize the vertical structure, where the proof mass is separated by a narrow air gap from a fixed plate, forming a parallel plate sense capacitance. In these devices, the proof mass moves in the direction peφendicular to its plane (z-axis) and changes the air gap. In a lateral accelerometer, a number of moving sense fingers are attached to the proof mass, and the sense capacitance is formed between these and the fixed fingers parallel to them. The sense direction in lateral accelerometers is in the proof-mass plane (x-y directions). Some designs use a "see-saw" structure, where a proof mass is suspended by torsional beams so that one side is heavier than the other side and in response to acceleration in the z- axis, the proof mas moves out of its plane. The advantages of this structure over conventional parallel-plate z-axis devices are built-in over range protection, larger sensitivity, and higher pull-in voltage.

The open-loop sensitivity of a capacitive accelerometer is proportional to the proof-mass size and capacitance overlap area and inversely proportional to the spring constant and air gap squared. Early micromachined capacitive accelerometers utilized bulk silicon micromachining and wafer bonding to achieve a thick, large proof mass and high sensitivity. One of the first reported devices used a silicon middle wafer anodically bonded to two glass wafers on top and bottom to form a z- axis accelerometer. The device had two differential sense capacitors, with the proof mass forming the middle electrode and metal on the glass wafers forming the top bottom fixed electrodes. The air gap was formed by recessing the silicon or glass wafers. This device with a proof-mass size of 4.6 mg and air gap of 2 μm provided μg-level performance. The second generation of this device had a resolution of better than 1 μg/VHz in a bandwidth of zero frequency to 100 Hz, with a temperature coefficient of offset (TCO) of 30 μg/°C and TCS of 150 ppm/°C. To reduce its temperature sensitivity and long-term drift, the later generation of this device was fabricated using three silicon wafers. Another significant early design with μg performance was fabricated using glass-silicon bonding and bulk micromachining and utilized a closed-loop ΣΔ readout and control circuit to achieve a 120 dB dynamic range.

Tunneling accelerometers: Some high-resolution physical sensors, including microaccelerometers, use a constant tunneling current between one tunneling tip (attached to a movable microstructure) and its counterelectrode to sense displacement. As the tip is brought sufficiently close to its counter-electrode (within a few angstroms) using electrostatic force generated by the bottom deflection electrode, a tunneling current ( tun) is established and remains constant if the tunneling voltage (Vtun) and distance between the tip and counterelectrode are unchanged. Once the proof mass is displaced due to acceleration, the readout circuit responds to the change of current and adjusts the bottom deflection voltage Vo to move the proof mass back to its original position, thus maintaining a constant tunneling current. Acceleration can be measured by reading out the bottom deflection voltage in this closed-loop system. Tunneling accelerometers can achieve very high sensitivity with a small size since the tunneling current is highly sensitive to displacement, typically changing by a factor of two for each angstrom of displacement. However, these devices have larger low- frequency noise levels.

Resonant accelerometers: The main advantage of resonant sensors is their direct digital output. The first resonant accelerometers were fabricated using quartz micromachining. Silicon resonant accelerometers are generally based on transferring the proof-mass inertial force to axial force on the resonant beams and hence shifting their frequency. To cancel device thermal mismatches and nonlinearities, a differential matched resonator configuration can be used. Recently, two high-sensitivity resonant accelerometers have been reported. These devices use wafer-thick proof mass and achieve high resolution (700 Hz/g with 524 kHz center frequency) and very good stability (2 μg in more than several days). However, these devices typically have small bandwidth (less than a few hertz). Also recently, surface-micromachined resonant accelerometers are developed. The resonator thereof consists of parallel beams, and its operation is based on rigidity change of the resonator due to its cross- sectional shape change, which is induced by the external acceleration.

Thermal accelerometers: Another class of accelerometers is based on thermal transduction. One of the first thermal accelerometers used the principle that the temperature flux from a heater to a heat-sink plate is inversely proportional to their separation. Hence, by measuring the temperature using thermopiles, the change in separation between the plates (which is representative of acceleration) can be measured. Devices with a moving thermopile array and fixed heater, and vice versa, can be fabricated. Recently, a novel thermal accelerometer was reported that does not have any moving mechanical parts. Its operation is based on free-convection heat transfer of a small hot air bubble in a sealed chamber. The device consists of a thermally isolated heater that forms a hot air bubble. The heat distribution of this bubble changes in the presence of an acceleration and becomes asymmetric with respect to the heater. This heat profile can be sensed by two symmetrically placed temperature sensors and is a measure of the acceleration.

Other accelerometers: In addition to the aforementioned accelerometers, accelerometers also use many other principles, including optical, electromagnetic, and piezoelectric. The motivation of the development of optical accelerometers has been combining optics and silicon micromachining to exploit advantages of both, as well as achieving miniature devices with very high EMI noise immunity [D. Uitamchandani, D. Liang, and B. Culshaw, "A micromachined silicon accelerometer with fiber optic integration" in Proc. SPIE Integrated Optics and Microstructures, 1992, pp. 27-33] or good linearity [R. S. Huang, E. Abbaspour-Sani, and C.

Y. Kwok, "A novel accelerometer using silicon micromachined cantilever supported optical grid and PIN photodetector", in Tech. Dig. 8 th Int. Conf. on Solid-State Sensors and Actuators (Transducers '95), Stockholm, Sweden, June 1995, pp. 663-666]. These devices, however, suffer two major limitations: (i) they relay on out-of-plane movement; and (ii) they fail to provide an "on-chip" light source and light sensor (e.g., photodiode) and therefore depend on a light source and detector which are off-chip.

Electromagnetic accelerometer reported typically utilizes two coils, one on top of the proof mass and the other separated by an air gap at the bottom, where the proof-mass displacement changes the mutual inductance of the two coils. By using a simple readout circuit a response is achieved.

Piezoelectric materials, mainly ZnO, have also been used in accelerometers to directly convert the force affecting the proof mass to an electrical signal. The piezoelectric charge generated by acceleration can be directly coupled to the gate of an MOS transistor and amplified. One of the problems with piezoelectric materials is their leakage that deteriorates the DC response of the device. Micromachined gyroscopes:

Almost all reported micromachined gyroscopes use vibrating mechanical elements to sense rotation. They have no rotating parts that require bearings, and hence they can be easily miniaturized and batch fabricated using micromachining techniques. All vibratory gyroscopes are based on the transfer of energy between two vibration modes of a structure caused by Coriolis acceleration. Coriolis acceleration, named after the French scientist and engineer G. G. de Coriolis, is an apparent acceleration that arises in a rotating reference frame and is proportional to the rate of rotation. To understand the Coriolis effect, imagine a particle traveling in space with a velocity vector v. An observer sitting on the x-axis of the xyz coordinate system, is watching this particle. If the coordinate system along with the observer starts rotating around the z-axis with an angular velocity Ω, the observer thinks that the particle is changing its trajectory toward the x-axis with an acceleration equal to 2vΩ. Although no real force has been exerted on the particle, to an observer, attached to the rotating reference frame an apparent force has resulted that is directly proportional to the rate of rotation. This effect is the basic operating principle underlying all vibratory structure gyroscopes. Resolution, drift, zero-rate output (ZRO), and scale factor are important factors that determine the performance of a gyroscope. In the absence of rotation, the output signal of a gyroscope is a random function that is the sum of white noise and a slowly varying function. The white noise defines the resolution of the sensor and is expressed in terms of the standard deviation of equivalent rotation rate per square root of bandwidth of detection [(°/s)/ ___z or (°/hNli_]. The so-called "angle random walk" in

Nh may be used instead. The peak-to-peak value of the slowly varying function defines the short- or long-term drift of the gyroscope and is usually expressed in °/s or °/h. Scale factor is defined as the amount of change in the output signal per unit change of rotation rate and is expressed in V/(°/s). Last, an important factor for any gyroscope that is primarily defined by device imbalances is the ZRO, which represents the output of the device in the absence of a rotation rate.

In general, gyroscopes can be classified into three different categories based on their performance: inertial-grade, tactical-grade, and rate-grade devices. Over the past few years, much of the effort in developing micromachined silicon gyroscopes has concentrated on "rate- grade" devices, primarily because of their use in automotive applications. This application requires a full-scale range of at least 50 °/s and a resolution of about OJ °/s in a bandwidth of 50 Hz. The operating temperature is in the range from -40 to 85 °C. There are also several other applications that require improved performance, including inertial navigation, guidance, robotics, and some consumer electronics. Today, optical gyroscopes are the most accurate gyroscopes available in the market. Among these, ring laser gyroscopes have demonstrated inertial-grade performance, while fiber-optic gyroscopes are mainly used in tactical-grade applications. Delco's hemispherical resonator gyroscope (HRG) is a vibratory gyroscope that has achieved impressive inertial-grade performance. Although highly accurate, these devices are too expensive and bulky for may low-cost applications. The reason for that is that these devices fail to provide (i) in-plane movement; and (ii) on-chip light source and sensor. Achieving "tactical- and inertial-grade" performance levels has proven to be a tough challenge for micromachined gyroscopes, and new technologies and approaches are being developed.

Micromachined vibratory gyroscopes:

A number of vibratory gyroscopes have been demonstrated, including tuning forks, vibrating beams, and vibrating shells. Tuning forks are a classical example of vibratory gyroscopes. The tuning fork gyroscope consists of two tines that are connected to a junction bar. In operation, the tines are differentially resonated to a fixed amplitude, and when rotated, Coriolis force causes a differential sinusoidal force to develop on the individual tines, orthogonal to the main vibration. This force is detected either as differential bending of the tuning fork tines or as a torsional vibration of the tuning fork the stem. The actuation mechanisms used for driving the vibrating structure into resonance are primarily electrostatic, electromagnetic, or piezoelectric. To sense the Coriolis-induced vibrations in the second mode, capacitive, piezoresistive, or piezoelectric detection mechanisms can be used. Optical detection is also feasible, but it is too expensive to implement in the prior art designs because such optical detection requires off-chip light source and sensor. In general, silicon micromachining processes for fabrication of vibratory gyroscopes fall into one of four categories: (i) silicon bulk micromachining and wafer bonding; (ii) polysilicon surface micromachining; (iii) metal electroforming and LIGA; and (iv) combined bulk-surface micromachining or so-called mixed processes. Piezoelectric vibratory gyroscopes were demonstrated in the early 1980's. Examples of these devices are fused quartz HRG by Delco, quartz tuning forks, like the Quartz Rate Sensor by Systron Donner, and a piezoelectric vibrating disc gyro. Although quartz vibratory gyroscopes can yield very high quality factors at atmospheric pressure with improved level of performance, their batch processing is not compatible with IC fabrication technology. In the late 1980's, after successful demonstration of batch- fabricated silicon accelerometers, some efforts were initiated to replace quartz with silicon in micromachined vibratory gyroscopes. The Charles Stark Draper Laboratory demonstrated one of the first batch fabricated silicon micromachined rate gyroscopes in 1991. This bulk silicon device was a double gimbal vibratoring mechanical element made from p++ silicon. The outer gimbal is electrostatically driven at a constant amplitude using the drive electrodes, and this oscillatory motion was transferred to the inner gimbal along the stiff axis of the inner flexures. When exposed to a rotation normal to the plane of the device, Coriolis force causes the inner gimbal to oscillate about its weak axis with a frequency equal to the drive frequency. Therefore, maximum resolution is obtained when the outer gimbal is driven at the resonant frequency of the inner gimbal. causing the sensitivity to be amplified by the mechanical quality factor of the sense resonance mode of the structure. A rotation rate resolution of 4/s in a 1 Hz bandwidth was realized using this structure.

Later in 1993, a 1 mm^ silicon-on-glass tuning fork gyroscope fabricated through the dissolved wafer process was reported. This gyroscope was electrostatically vibrated in its plane using a set of interdigitated comb drives to achieve a large amplitude of motion (10 um). Any rotation in the plane of the substrate peφendicular to the drive mode will then excite the out-of-plane rocking mode of the structure, which is capacitively monitored. Other tuning-fork designs have used electromagnetic excitation to obtain a large amplitude of motion. Bosch's silicon yaw rate sensor achieves vibration amplitudes as large as 50 μm using a permanent magnet mounted inside a metal package. This device was fabricated through a combination of bulk- and surface- micromachining processes, and it consists of two bulk micromachined oscillating masses, each of which supports two surface-micromachined accelerometers for detecting of Coriolis force. The sensor chip is anodically bonded to a supporting glass wafer and is covered by another silicon cap wafer. Operating at atmospheric pressure, the device has shown a resolution of 0.3/5 in a 100 Hz bandwidth, thanks to its large amplitude of vibration. Although such a large amplitude of oscillation (50 μm) can increase the output signal level, it increases the total power consumption and may cause fatigue problems over long-term operation. Cross talk-between the sense and drive modes was minimized through mechanical decoupling of these modes by separating the oscillator and sense proof masses, resulting in a stable ZRO.

Piezoresistive detection has also been used in some gyroscope designs. Daimler Benz has demonstrated a tuning fork angular rate sensor for automotive applications that piezoresistively measures the rotation- induced shear stress in the stem of the tuning-fork device. In this device, a piezoelectric aluminum nitride (A1N) thinfilm layer on one of the times. The use of piezoelectric thin films such as A1N and ZnO on silicon degrades Q and causes large temperature variation of offset and sensitivity. This device was fabricated through a combination of bulk micromachining and bonding of SOI wafers. Researchers at the University of Neuchatel, Switzerland, have demonstrated a tuning-fork design based on two isolated vibrating proof masses, each supported by a four beam bridge suspension. These proof masses are electromagnetically vibrated in plane and antiphase, and the rotation-induced out-of-plane motion is then detected by means of four piezoresistors connected in a Wheatstone bridge configuration, showing a sensitivity of 4 nV/% with excellent linearity up to 750/5. This device was fabricated through silicon bulk micromachining and was wafer- level vacuum packaged by anodic bonding of the silicon wafer to encapsultaing glass wafers. In general, package-induced stress on the sensor structure can be lowered by low-temperature anodic bonding of glass wafers with silicon. Although piezoresistive devices are easier to fabricate and require a simpler electronic interface due to their lower output impedance compared to capacitive devices, they have large temperature sensitivity and poor resolution.

Also reported in the literature are capacitive bulk micromachined silicon-on-glass vibrating beams, vibrating membranes, and double- gimbaled structures. Since the Young's modulus of single-crystal silicon changes with crystallographic orientation, symmetric vibrating structures made of single-crystal silicon may show excessive mechanical coupling between drive and sense modes (due to this anisotropy), resulting in a large ZRO with unacceptable drift characteristics. Surface-micromachined vibratory gyroscopes have also been demonstrated. Some have been integrated with the readout electronic circuitry on a single silicon chip, reducing parasitic capacitances and hence increasing the signal-to-noise ratio. In addition, the vibrating structure is made of polysilicon, which has a high quality factor and an orientation- independent Young's modulus. Single and dual-axis polysilicon surface- micromachined gyroscopes have been realized by researchers at Berkeley and Samsung. Berkeley's z-axis vibratory rate gyroscope resembles a vibrating beam design and consists of an oscillating mass that is electrostatically driven into resonance using comb drives. Any deflections that result from Coriolis acceleration are detected differentially in the sense mode using interdigitated comb fingers. This device, 1 mm across, was integrated with a transresistance amplifier on a single die using the Analog Devices BiMEMS process. The remaining control and signal-processing electronics were implemented off-chip. Quadrature error nulling and sense- mode resonant frequency tuning can be accomplished in this design by applying a control DC bias voltage to the position sense fingers. The DC bias voltage generates an electrostatic negative stiffness, which reduces the resonant frequency of the sense mode. By slightly changing this DC bias voltage on the differential comb fingers (+ ΔV), a lateral electrostatic field arises that can be used to align the drive mode oscillations and reduce the quadrature error. Samsung has also reported a very similar surface- micromachined z-axis device. In this case, hybrid attachment of the sensor chip to a CMOS application specific integrated circuit (ASIC) chip used for readout and closed-loop operation of the gyro was done in a vacuum- packaged ceramic case.

Murata has presented a surface-micromachined polysilicon gyroscope that is sensitive to lateral (x- or j-axis) angular rate. The sense electrode was made underneath the perforated polysilicon resonator by diffusing phosphorous into the silicon substrate (junction isolation). The junction-isolation scheme used in this device, although simple, has the disadvantage or relatively large parasitic capacitance and large amount of shot noise associated with the existing pn junction, which in turn degrade the resolution. Later in 1997, Samsung reported a similar device that used a 3000

Angstrom thick polysilicon sense electrode underneath a 705 μm-thick low- pressure chemical vapor deposition (LPCVD) polysilicon resonating mass. Since the detection mode is highly damped by squeeze film damping, these devices have to operate under vacuum. Samsung's device, vacuum packaged in an AI2O3 case, showed an improved open-loop noise- equivalent rate.

Berkeley has reported a surface-micromachined dual-axis gyroscope based on rotational resonance of a 2-μm-thick polysilicon rotor disk. Since the disk is symmetric in two orthogonal axes, the sensor can sense rotation equally about these two axes. This device integrated with electronics, yielded a random walk as low as 10° 'Nh with cross-axis sensitivity ranging 3-16 %. Resolution can be further improved to 2°Nh by frequency matching at the cost of excessive cross-axis sensitivity. Also reported in the literature is a cross-shaped nickel-on-glass two axis micromachined gyroscope, which has shown a rate sensitivity of 0.1 __V/°/s.

The JPL in collaboration with the University of California, Los Angles has demonstrated a bulk-micromachined, precision silicon MEMS vibratory gyroscope for space applications. This clover-leaf shaped gyroscope consists of three major components: a silicon clover-leaf vibrating structure; a silicon baseplate, which is bonded to the clover-leaf structure; and a metal post, which is epocied inside a hole on the silicon resonator. A hermetically sealed package houses the microgyroscope and most of its control electronics.

Recently, researchers at HSG-IMIT, Germany, have demonstrated and reported a surface-micromachined precision x-axis vibratory gyroscope (MARS-RR) with a very small ZRO achieved by mechanical decoupling of the drive and sense vibration modes. This device was fabricated through the standard Bosch foundry process featuring a 10-μm-thick structural polysilicon layer in addition to the buried polysilicon layer, which defines the sense electrodes.

Researches at General Motors and the University of Michigan have developed a vibrating ring gyroscope which consists of a ring, semicircular support springs, and drive, sense, and balance electrodes, which are located around the structure. Symmetry considerations require at least eight springs to result in a balanced device with two identical flexural modes that have equal natural frequencies. The ring is electrostatically vibrated into an in- plane elliptically shaped primary flexural mode with a fixed amplitude. When it is subjected to rotation around its normal axis, Coriolis force causes energy to be transferred from the primary mode to the secondary flexural mode, which is located 45° apart from the primary mode, causing amplitude to build up proportionally in the latter mode; this buildup is capacitively monitored. The vibrating ring structure has some important features compared to other types of vibratory gyroscopes. First, the inherent symmetry of the structure makes it less sensitive to spurious response. Second, since two identical flexural modes of the structure "with nominally equal resonant frequencies" are used to sense rotation, the sensitivity of the sensor is amplified by the quality factor of the structure, resulting in higher sensitivity. Third, the vibrating ring is less temperature sensitive since the vibration modes are effected equally by temperature. Last, electronic balancing of the structure is possible. Any frequency mismatch due to mass or stiffness asymmetries that occurs during the fabrication process can be electronically compensated by use of the balancing electrodes that are located around the structure.

The first micromachined version of the vibrating ring gyroscope was fabricated by electroforming nickel into a thick polyimide (or photoresist) mold on a silicon substrate in a post circuit process. The gyroscope demonstrated a resolution of 0.5 -? in a 25 Hz bandwidth limited by the readout electronic noise. The sensor was integrated with a low-input capacitance source-follower buffer and the amplifier on a silicon chip. The zero bias drift was <10/5 over the temperature range of -40 to 85 °C, and the sensitivity of the device varied by less than 3 % over the same temperature range.

To improve performance further, a new polysilicon ring gyroscope (PRG) was recently fabricated through a single-wafer, all-silicon, high- aspect-ratio p++/polysilicon trench-refill technology at the University of Michigan. In this new process, the vibrating ring and support springs are created by refilling deep dry-etched trenches with polysilicon deposited over a sacrificial LPCVD oxide layer. Each sense electrode is made from a p++ silicon island (12 μm deep) hanging over an ethylenediamine-pyrocatechol (EDP)-etched pit. This device provides several important features required for high-performance gyroscopes, including small ring-to-electrode gap spacing (<1 μm) for increasing the sense capacitance; large structural height for increasing the radius and sense capacitance; large structural height for increasing the radius and sense capacitance and reducing the resonant frequency; and a better structural material (polysilicon) for increasing Q with an orientation-independent Young's modulus.

British Aerospace Systems and Equipment, in collaboration with Sumitomo Precision Products, has also developed a micromachined single- crystalline silicon ring gyroscope with a reported root-mean-square noise floor of 0J5/5 at a 30 Hz bandwidth and a in-run drift of approximately 0.05/5. This device was fabricated through deep dry etching of a 100-μm- thick silicon wafer, which was then anodically bonded to a glass support wafer. Levitated micromachined spinning-disc gyroscopes have also been investigated. The concept was based on a rotor disc, levitated using electromagnetic or electrostatic means and spun at a very high rate by means of a motor to produce angular momentum. With additional electrostatic fields, the rotor can be held in equilibrium even if the sensor is tilted or inverted. It is predicted that spinning microgyroscopes can yield a lower drift than a vibrating structure gyroscope. The performance of these devices is yet to be demonstrated.

It will be appreciated that although its highest precision, most the above described inertial sensors fail to employ optical detection of inertial imposed movement. The reason for that relies on (i) the present need to use off-chip light source and sensor; and (ii) the high cost of such off-chip light source and sensor. In addition, the described inertial sensors described fail to employ an in-plane (x-y) inertial movement combined with an on-chip detection of inertial. There is thus a widely recognized need for, and it would be highly advantageous to have, micromachined inertial sensors employing in-plane (x-y) degree of freedom combined with on-chip detection of inertial.

SUMMARY OF THE INVENTION According to one aspect of the present invention there is provided a micromachined chip comprising (a) a reference frame; (b) at least one suspended waveguide element having an in-plane degree of freedom being integrally formed with the reference frame; and (c) a light source being integrally formed with the reference frame and being optically coupled to the at least one suspended waveguide element at one end thereof.

According to another aspect of the present invention there is provided a micromachined chip comprising (a) a reference frame; (b) at least one suspended waveguide element having an in-plane degree of freedom being integrally formed with the reference frame; and (c) a light sensor being integrally formed with the reference frame and being optically coupled to the at least one suspended waveguide element at one end thereof.

According to yet another aspect of the present invention there is provided a micromachined displacement sensor chip comprising (a) a reference frame; (b) at least one suspended waveguide element having an in-plane degree of freedom being integrally formed with the reference frame; (c) a light source being integrally formed with the reference frame and being optically coupled to the at least one suspended waveguide element at one end thereof; and (d) a light sensor being integrally formed with the reference frame and being optically coupled to the at least one suspended waveguide element at another end thereof; such that when the at least one suspended waveguide element is subjected to an external force, an in-plane displacement of the at least one suspended waveguide element is monitorable by the light sensor due to light modulation.

According to an additional aspect of the present invention there is provided a micromachined displacement sensor chip comprising (a) a reference frame; (b) at least one suspended element having an in-plane degree of freedom being integrally formed with the reference frame through a root thereof; (c) a solid state sensor being integrally formed in the root, such that when the at least one suspended element is subjected to an external force, an in-plane displacement of the at least one suspended element is monitorable by the solid state sensor.

According to further features in preferred embodiments of the invention described below, the micromachined displacement sensor chip further comprising at least one suspended proof mass integrally formed with the at least one suspended waveguide element.

According to still further features in the described preferred embodiments the at least one suspended waveguide element also serves as a suspended proof mass .

According to still further features in the described preferred embodiments the micromachined displacement sensor chip further comprising at least one fixed waveguide element integrally formed with the reference frame, the at least one fixed waveguide is optically coupled through one end thereof to the light source and through another end thereof to the at least one suspended waveguide element, thereby optically coupling between the at least one suspended waveguide element and the light source.

According to still further features in the described preferred embodiments the at least one fixed waveguide element includes a splitting fixed waveguide element and further wherein the at least one suspended waveguide element includes a combining suspended waveguide element, the splitting fixed waveguide element and the combining suspended waveguide element are optically coupled such that light arriving from the light source and guided through the splitting fixed waveguide element recombines in the combining suspended waveguide element to thereby form an interferometer.

According to still further features in the described preferred embodiments the micromachined displacement sensor chip further comprising at least one fixed waveguide element integrally formed with the reference frame, the at least one fixed waveguide is optically coupled through one end thereof to the light sensor and through another end thereof to the at least one suspended waveguide element, thereby optically coupling between the at least one suspended waveguide element and the light sensor. According to still further features in the described preferred embodiments the at least one fixed waveguide element and the at least one suspended waveguide element are optically coupled via a tip-tip optical coupling. According to still further features in the described preferred embodiments the at least one fixed waveguide element and the at least one suspended waveguide element are optically coupled via a tip-blunt end optical coupling.

According to still further features in the described preferred embodiments the at least one fixed waveguide element and the at least one suspended waveguide element are optically coupled via a blunt end-tip optical coupling.

According to still further features in the described preferred embodiments the at least one fixed waveguide element and the at least one suspended waveguide element are optically coupled via a reflector to effect geometrical modulation in reflection mode.

According to still further features in the described preferred embodiments the force is an acceleration force, the micromachined displacement sensor chip serves as a micromachined accelerometer chip. According to still further features in the described preferred embodiments the force is a Coriolis force, the micromachined displacement sensor chip serves as a micromachined gyroscope chip.

According to still further features in the described preferred embodiments the force is a Coriolis force, the micromachined displacement sensor chip further includes an electrostatic actuator integrally formed with the reference frame for actuating the proof mass in plane vertically to the

Coriolis force, the micromachined displacement sensor chip serves as a micromachined gyroscope chip. According to still further features in the described preferred embodiments the force is selected from the group consisting of thermal expansion force, electrostatic force, magnetic force and piezoelectric force, whereas the suspended waveguide element is selected responsive to the thermal expansion force, the electrostatic force, the magnetic force and the piezoelectric force, respectively. Thus, the micromachined displacement sensor chip can serve as a sensor for such forces or as a sensor for a microactuator actuated in response to such forces.

According to still another aspect of the present invention there is provided a micromachined displacement sensor chip comprising (a) a reference frame;

(b) a first waveguide element being integrally formed with the reference frame; (c) a second waveguide element being integrally formed with the reference frame; (e) a light source being integrally formed with the reference frame and being optically coupled to the first waveguide element at one end thereof; and (f) a light sensor being integrally formed with the reference frame and being optically coupled to the second waveguide element at one end thereof; (g) a reflector integrally formed with the reference frame and optically coupling the first waveguide element with the second waveguide element; wherein at least one of the reflector, the first waveguide element and the second waveguide element serves as a suspended proof mass to effect geometrical modulation in reflection mode when displaced as a response to an external force exerted thereon.

The present invention successfully addresses the shortcomings of the presently known configurations by providing displacement sensor and actuator chips which include both a force responding mechanism and integral displacement sensing mechanism.

BRIEF DESCRIPTION OF THE DRAWINGS The invention is herein described, by way of example only, with reference to the accompanying drawings, wherein:

FIGs. la-b schematically and photographically, respectively, present a single crystal silicon (SCS) beam which serves as a suspended optical waveguide (SOW) and can be used in the construction of an inertial sensor according to the present invention.

FIGs. 2a-c show photographs (Figures 2b-c) of light emitted form a SOW which is shown in Figure 2a. FIG. 3 shows a simplest displacement sensing application employing such a SOW and a suspended proof mass connected thereto.

FIGs. 4-5 schematically and photographically, respectively, show a displacement sensing application employing a planar interferometer based displacement sensor.

FIG. 6 is a graph showing the response of the sensor of Figure 5 to shock and reveals the resonance response of the sensor.

FIG. 7 is a graph showing the calculated sensitivity of the sensor of Figure 5 and compares it to the sensitivity of the simple sensor shown in Figure 3 in far field.

FIG. 8 is a scanning electron microscope image showing a prior art Siθ2-Al2θ3-Siθ2 waveguide with out-of-plane degree of freedom (DOF).

FIG. 9 is a schematic presentation of a the integration of optical links with fixed and suspended waveguides which is referred to herein as integrated SOW or /-SOW, according to the present invention.

FIGs. lOa-b show a LED waveguide direct butt coupling and a LED- waveguide butt coupling with adiabatic mode converter (AMC).

FIG. 11 is a schematic depiction of a light emitting polymer (LEP) integrated at the end of a Si3N-ι waveguide according to the present invention.

FIG. 12 is a schematic view of a SOW building block.

FIG. 13 is a cross sectional view of the SOW building block of Figure 12.

FIG. 14 shows the guiding parameter of a 10 μm tall silicon waveguide as a function of its width for λ=0.63 μm, and for the parameters shown in Figure 13.

FIG. 15 show the conditions in Equations 7 and 8 as a function of the silicon waveguide width, for the parameters used in Figure 14. The parameters are drawn for p=T, q=l, and for p=3, q=3. For a=1.6 μm the conditions set by Equations 7 and 8 hold.

FIG. 16 shows the guiding parameter of a 1.6 μm wide silicon nitride waveguide as a function of its thickness, for λ=1.3 μm, and for the parameters shown in Figure 13.

FIGs. 17a-b are schematic depictions of a fixed-free suspended waveguide, (a), and the modes in the different waveguides (b).

FIG. 18 shows the coupling between the first mode in the sending waveguide and the modes in the receiving waveguide, as a function of the displacement. The graph is drawn for the suspended silicon waveguide with the following parameters: a=1.6 μm, b=10 μm, Z =2 μm, and λ=1.3 μm.

FIG. 19 shows the sensitivity of the coupling as a function of the displacement. The graph is drawn for the same parameters as in Figure 18. The sensitivity of the first modes is maximum for an initial displacement of half of the waveguide width.

FIG. 20 shows the relative sensitivity as a function of the displacement. The relative sensitivity of the coupling with the first mode, (Ci _j), is linear with the displacement, but deviates from linearity for the higher symmetrical modes (i.e., CR-13, CR-1 5). CR-12, CR1 on the other hand overlap at small displacement, peak at about δ=0J μm, and decreases as the displacement increases. At large displacement

and

split and merge with the relative sensitivities of the symmetric modes.

FIGs. 21a-b are schematic depictions of sensing configuration based on Tip- Waveguide interaction and tip-tip interaction.

FIGs. 22a-h are schematic views of a SCREAM process used to fabricate a SOW.

FIGs. 23a-b are top and side views of a testing setup for mechanical excitation. FIG. 24 is an scanning electron SEM picture showing the general view of a chip containing several sensors in a fixed-fixed configuration. The sensors are fabricated on top of a plateau formed by KOH etching.

FIG. 25 is a scanning electron micrograph of a fixed- free SOW with a proof mass at its end. FIG. 26 is a close view showing the parabolic horn expansion and the two blocking planes placed vertically to the waveguide.

FIG. 27 is a close view of the tip area of the fixed-free SOW, showing an initial misalignment.

FIG. 28 shows the response of the fixed-free SOW displacement sensor. The upper curve is the vibrational frequency response when the device is positioned below the optical path. This spectrum contains a noisy DC signal resulting from the laser light collected by the photodiode. As the sensor, aligned to the light path, is raised up and blocks the light path the DC signal drops and a resonance signal appears as shown by the lower curve.

FIG. 29 shows the response of the device in Figure 26 to a mechanical white noise. FIG. 30 is a scanning electron micrograph of a fixed-fixed version of a SOW.

FIG. 31 shows the response of the device in Figure 30 to acoustical excitation. FIG. 32 is a scanning electron micrograph showing a micro platform suspended over four bending beam. The suspending beams serve also as suspended waveguide with GM configuration.

FIG. 33 is a close view of the tip area of the SOW integrated micro- platform shown in Figure 30. FIG. 34 is a schematic view of the SOW building block according to one embodiment of the present invention.

FIG. 35 is a general view of a GM-RM sensor according to the present invention.

FIG. 36 is a close view of the input-output waveguides of the GM- RM sensor according to the present invention.

FIGs. 37a-b are schematic side and top view showing the plateau where the GM-RM sensor is fabricated.

FIG. 38 is a schematic view of the testing setup used to test the GM- RM sensor of Figures 35-36. FIG. 39 shows the time response of the GM-RM sensor of Figures

35-36 to a pulse excitation.

FIG. 40 show the spectral response of the GM-RM sensor of Figures 35-36 to a pulse excitation.

FIG. 41 shows the first four vibration modes of the suspended structure of the GM-RM sensor of Figures 35-36 as calculated using FEA. The dashed line represent the un-deformed state of the suspended structure.

FIG. 42 is a schematic representation of an GMI displacement sensor. Light is divided between two fixed waveguides aligned to suspended waveguides at distances z_a, and ztø. By modulation z_a, and ztø the light waves enter the suspended waveguides with a different phase. This phase is translated into intensity at the merged section of the suspended waveguides.

FIG. 43 shows the output intensity as a function of the displacement for two configurations: a symmetric configuration with z_a=zb=2 μm and an asymmetric configuration with z_a=1.7 μm and z_D=2 μm. For comparison, the output intensity resulting from the coupling efficiency of the first mode (Cj 1 ), as a function of the displacement in a GM displacement sensor is drawn (dashed line). FIG. 44 shows the sensitivity of the GMI displacement sensor as a function of the displacement. The sensitivity is calculated for z_aN2 μm and Ztø=L68 μm. For comparison, the output intensity as a function of the displacement for the first mode in a GM displacement sensor is drawn (dashed line). Both the configurations of the GMI and the GM, are designed such that the zero displacement sensitivity is maximized.

FIG. 45 is a scanning electron micrograph showing the general view of a chip containing several sensors. The sensors are fabricated on top of a plateau formed by KOH etching. FIGs. 46a-b are general views of GMI based on the concept depicted in Figure 42.

FIGs. 47 and 48 are enlarged section of Figure 46 showing the fixed and free waveguides and the merged section of the suspended waveguide that feeds the optical sensor through a 5 μm wide fixed waveguide. FIG. 49 is a schematic view of the testing setup.

FIG. 50 shows the response of the GMI displacement sensor to acoustical excitation in the 2-6 kHz range.

FIG. 51 is a Finite Element Analysis model of the GMI. FIG. 52 shows the response of the GMI to an impulse excitation. FIG. 53 is a schematic representation showing a tunable GMI displacement sensor. The comb drive drives the mesh to the working position.

FIG. 54 shows the calculated sensitivity of the tunable GMI displacement sensor as a function of z_a for z_a+z_D=6 μm. FIG. 55 is a schematic depiction of a sensor a placement of a solid state sensor relative to suspended silicon beam according to the present invention.

FIG. 56a is a canning electron micrograph (SEM) of two pn diodes integrated at the root of a single silicon beam. FIG. 56b is a scanning electron micrograph of an NMOS transistor integrated at the root of a single silicon beam.

FIG. 56c is a scanning electron micrograph of two NMOS transistors with a common source, integrated at the root of double parallel silicon beams. FIG. 56d is a scanning electron micrograph of two NMOS transistors with a common source integrated at the root of double vertical silicon beams. FIG. 57a is a scanning electron micrograph of a meshed proof mass supported by four beams. FIG. 57b is a close view of one of the supporting beams of the proof mass of Figure 57a showing an integrated NMOS transistor in more detail. FIG. 57c shows the frequency response of the sensor shown in

Figure 57a.

FIG. 57d shows the response of the sensor in Figure 57a to acceleration.

FIG. 58a is a scanning electron micrograph of a sensor comprising a proof mass supported by one L shaped beam where two diodes in the configuration shown in Figure 56a are integrated at the root of the beam.

FIG. 58b shows the amplified response of the sensor in Figure 58a to acceleration of 0.03 g at different frequencies and different pressure level.

FIG. 58c shows the amplified response of the sensor shown in Figure 58a to acceleration with different amplitudes at 1 kHz, 4 kHz and at resonance (4.3 kHz).

FIG. 58d shows the amplified response of the sensor in Figure 58a to acoustical and mechanical shocks.

FIG. 59a is a scanning electron micrograph of a cantilevered coiled beam with proof mass at its end and a single NMOS transistor at its root.

FIG. 59b is a closer view of the integration of the NMOS transistor with the silicon beam shown in Figure 59a.

FIG. 59c shows the amplified response of the sensor shown in Figure 59a to acceleration of 1.7 g at pressure of 2 mbar, from four directions. FIG. 59d shows the amplified response of the sensor shown in Figure

59a to flow of nitrogen at different flow rates, showing both amplitude and frequency dependence over the flow rate.

FIG. 60a is a scanning electron micrograph of a comb drive actuator comprising a meshed mass supported by four bending beams. FIG. 60b is a close view of the actuator shown in Figure 60a showing the integration of an NMOS transistor with one of the four supporting beams.

FIG. 61 is a schematic depiction showing the effect of Coriolis force on the rotation of a ballerina. FIG. 62 is a schematic depiction shows how this effect is used to sense the rotation rate in a conventional (prior art) vibrating gyroscope. FIGs. 63-64 are schematic depictions of a gyroscope according to one aspect of the present invention, which gyroscope employs integrated light source and light sensor for sensing the Coriolis force.

FIG. 65 is a schematic depiction of a gyroscope according to another aspect of the present invention, which gyroscope employs an integrated solid state sensor for sensing the Coriolis force.

FIG. 66 is a schematic depiction of a micromachined chip according to one aspect of the present invention which incoφorates a suspended waveguide capable of in plane displacement, an integral light source and optionally or alternatively an integral light sensor.

FIG. 67 is a schematic depiction of a micromachined displacement sensor chip according to another aspect of the present invention which includes at least one suspended waveguide element having an in-plane (x-y) degree of freedom, a light source and a light sensor arranged such that when the element is subjected to an external force, an in-plane displacement thereof is monitorable by the light sensor due to light modulation.

FIG. 68 is a schematic depiction of tip-tip optical coupling according to the present invention.

FIG. 69 is a schematic depiction of tip-blunt end optical coupling according to the present invention.

FIG. 70 is a schematic depiction of blunt end-tip optical coupling according to the present invention.

FIG. 71 is a schematic depiction of a fixed waveguide element and a suspended waveguide element optically coupled via a reflector to effect geometrical modulation in reflection mode according to the present invention.

FIG. 72 is a schematic depiction a splitting fixed waveguide element and a combining suspended waveguide element arranged along with a light source and a light sensor into an interferometer. FIG. 73 is a schematic depiction of a gyroscope responding to a

Coriolis force according to the present invention.

FIG. 74 is a schematic depiction of a micromachined displacement sensor incoφorating a first waveguide element, a second waveguide element, a light source, a light sensor and a reflector optically coupling the first waveguide element with the second waveguide element, whereas at least one of the reflector, first waveguide element and second waveguide element serves as a suspended proof mass to effect geometrical modulation in reflection mode when displaced as a response to an external force exerted thereon.

FIG. 75 is a schematic depiction of a micromachined displacement sensor chip which includes at least one suspended element having an in- plane degree of freedom and which is integrally formed with a reference frame through a root thereof which incoφorates a solid state sensor, such that when the suspended element is subjected to an external force, an in- plane displacement thereof is monitorable by the solid state sensor.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

The present invention is of micromachined displacement sensors and actuators which employ on chip light source and/or light sensor, which can be used to monitor minute displacements due to an external force such as a gravitational force. Specifically, the present invention can be used to fabricate highly sensitive, accurate, yet low cost accelerometer and gyroscope sensors, as well as actuators which are responsive to a variety of forces. Most, specifically, the present invention provides such sensors and actuators employing in-plane degree of freedom combined with on-chip detection of displacement. The principles and operation of a sensor or actuator according to the present invention may be better understood with reference to the drawings and accompanying descriptions.

Before explaining at least one embodiment of the invention in detail, it is to be understood that the invention is not limited in its application to the details of construction and the arrangement of the components set forth in the following description or illustrated in the drawings. The invention is capable of other embodiments or of being practiced or carried out in various ways. Also, it is to be understood that the phraseology and terminology employed herein is for the puφose of description and should not be regarded as limiting.

Referring now to the drawings, Figures 66-75 illustrate several embodiments of the present invention.

As shown in Figure 66, according to one aspect of the present invention a micromachined chip 100 is provided. Chip 100 includes a reference frame 102, at least one suspended waveguide element 104 having an in-plane (x-y) degree of freedom and which is integrally formed with reference frame 102, and a light source 106 which is integrally formed with reference frame 102 and which is optically coupled to suspended waveguide element 104 at one end thereof.

As further shown in Figure 66, according to another aspect of the present invention chip 100 alternatively or additionally includes a light sensor 108 which is integrally formed with reference frame 102 and is optically coupled to suspended waveguide element 104 at one end thereof. In a configuration wherein chip 100 includes both light source 106 and light sensor 108, chip 100 can serve as a simple accelerometer in which acceleration force acting on element 104 and is monitorable by sensor 108, because the degree to which element 104 displaces as a result of the force exerted thereon and therefore the degree of light modulation sensed by sensor 108 correlates to the degree of acceleration force.

For details relating to the integration of light sources and light sensors in a chip and for details relating to the formation of a suspended waveguide having an in-plane degree of freedom, see the Examples section that follows.

Another aspect of the present invention is shown in Figure 67. A micromachined displacement sensor chip 110 is provided and includes a reference frame 112, at least one suspended waveguide element 114 having an in-plane (x-y) degree of freedom and which is integrally formed with reference frame 112, a light source 116 integrally formed with reference frame 112 and which is optically coupled to element(s) 114 at one end thereof, and a light sensor 118 integrally formed with reference frame 102 and which is optically coupled to element(s) 114 at another end thereof, such that when element 114 is subjected to an external force, an in-plane (x- y) displacement thereof is monitorable by light sensor 118 due to light modulation.

According to one embodiment of the present invention element 114 also serves as a suspended proof mass. According to another embodiment of the present invention chip 110 further includes at least one suspended proof mass 120 which is integrally formed with waveguide element(s) 114. As further shown in Figure 67, according to a preferred embodiment of the present invention chip 110 further includes at least one fixed waveguide element 122 integrally formed with reference frame 112 and which is optically coupled through one end thereof to light source 116 and through another end thereof to suspended waveguide element(s) 114, thereby optically coupling between element(s) 114 and light source 116. As still further shown in Figure 67, according to a preferred embodiment of the present invention chip 110 further includes chip 100 further includes at least one fixed waveguide element 124 integrally formed with reference frame 112 and which is optically coupled through one end thereof to light sensor 118 and through another end thereof to suspended waveguide element(s) 114, thereby optically coupling between element(s) 114 and light sensor 118.

As shown in Figure 68, according to one embodiment of the present invention fixed waveguide element(s) 122/124 and suspended waveguide element(s) 114 are optically coupled via a tip-tip optical coupling.

As shown in Figure 69, according to another embodiment of the present invention fixed waveguide element(s) 122/124 and suspended waveguide element(s) 114 are optically coupled via a tip-blunt end optical coupling. As shown in Figure 70, according to still another embodiment of the present invention fixed waveguide element(s) 122/124 and suspended waveguide element(s) 114 are optically coupled via a blunt end-tip optical coupling.

As shown in Figure 71, according to yet another embodiment of the present invention fixed waveguide element(s) 122/124 and suspended waveguide element(s) 114 are optically coupled via a reflector 126, integrally formed with reference frame 112, to effect geometrical modulation in reflection mode.

As shown in Figure 72, according to yet another embodiment of the present invention fixed waveguide element(s) 122 includes a splitting fixed waveguide element 122' and further wherein suspended waveguide element(s) 114 includes a combining suspended waveguide element 114'. Splitting fixed waveguide element 122' and combining suspended waveguide element 114' are optically coupled such that light arriving from light source 116 and guided through splitting fixed waveguide element 122' recombines in combining suspended waveguide element 114' to thereby form, along with detector 118, an interferometer.

As shown in Figure 73, according to yet another embodiment of the present invention chip 110 is a gyroscope responding to a Coriolis force indicated by arrow 132 and which is effected by rotation as indicated by arrow 134. Chip 100 further includes an electrostatic actuator 130 integrally formed with reference frame 112 for actuating proof mass 120, as indicated by arrow 136 in plane, vertically to the Coriolis force. As shown in Figure 74, according to yet another embodiment of the present invention there is provided a micromachined displacement sensor chip 140 which includes, a reference frame 142, a first waveguide element 144 integrally formed with reference frame 140, a second waveguide element 145 integrally formed with reference frame 142, a light source 146 integrally formed with reference frame 142 and which is optically coupled to first waveguide element at one end thereof, a light sensor 148 integrally formed with reference frame 142 and which is optically coupled to second waveguide element 145 at one end thereof, and a reflector 149 integrally formed with reference frame 142 and optically coupling first waveguide element with second waveguide element. At least one of reflector 149, first waveguide element 144 and second waveguide element 145 serves as a suspended proof mass to effect geometrical modulation in reflection mode when displaced as a response to an external force exerted thereon. As shown in Figure 75, according to still another aspect of the present invention there is provided a micromachined displacement sensor chip 160 which includes a reference frame 162, at least one suspended element 164 having an in-plane degree of freedom and which is integrally formed with reference frame 162 through a root 165 thereof. Chip 160 further includes a solid state sensor 166 which is integrally formed in root

165, such that when suspended element(s) 164 is subjected to an external force, an in-plane displacement thereof is monitorable by solid state sensor

166. Further detailed relating to the functionality of the solid state sensors are provided in the Examples section that follows. Each one of the sensor chips described herein can serve as an actual sensor to sense and report a magnitude such as an inertial magnitude. However, each one of the sensor chips can alternatively serves as a sensor for actuating a microactuator. As such, any force which is used for microactuation, such as, but not limited to, thermal expansion force, electrostatic force, magnetic force or piezoelectric force can be employed, provided that the suspended element is selected responsive to such force. Yet alternatively, such forces can be senses by the sensor and thereby be monitorable for non-microactuation puφoses.

Additional objects, advantages, and novel features of the present invention will become apparent to one ordinarily skilled in the art upon examination of the following examples, which are not intended to be limiting. Additionally, each of the various embodiments and aspects of the present invention as delineated hereinabove and as claimed in the claims section below finds experimental support in the following examples.

EXAMPLES

5 Reference is now made to the following examples, which together with the above descriptions, illustrate the invention in a non limiting fashion.

EXAMPLE 1 l o Technological feasibility of silicon micromachined inertial sensors employing integrated optical sensing

This Example describes a monolithic integration of two types of technologies on a single silicon chip which can be used to allow breaking through abilities in several fields including in the field of micromachined

15 inertial sensors such as accelerometers and gyroscopes.

The integration includes an optical link (a light source, a waveguide, and a photodetector), in which the trarismission medium (the waveguide) is a micro-opto-electro-mechanical sensor, based on a suspended optical waveguide (SOW) technology. This technology implements highly 0 sensitive sensors that are based on near field, interferometery, and evanescent field.

Two types of light sources are described herein. One is based on III- V-Nitride Semiconductors deposited monolithically on an silicon wafer, and the other is based on polymers light emitting diodes (LED). In addition to 5 the light sources an integration with an optical detector is further described. Silicon based photodiodes are currently available and therefore this integration is much easier.

It will be appreciated that optical sensors without such integration are accessed through optical fibers that inflict several drawbacks: (i) it requires

30 hand work and thus has low yield and high cost; (ii) it increases the noise level by loosing information through the free-space interconnect; and (iii) It requires hybrid interface to very large scale integration (VLSI) and thus is associated with relatively low signal to noise ratio.

As further detailed hereinunder, the integration described herein

35 allows on-chip optical manipulations that are accessed and controlled via on chip electronics. In addition, this integration allows to further process the optical signal on the silicon chip using IC, and thus will increase the speed of processing and signal to noise ratio. Lastly, as the microfabrication technology is based on VLSI technology, mass production becomes an inherent property of this technology.

Optical sensors of mechanical effects are among the most sensitive known [Sze, 1990]. For example optical manipulation in near field is used to reconstruct surfaces down to atomic level [Pohl, 1990; Betzig and Trauman, 1992]. Using interferometric means, displacement sensors with Angstrom resolution can be configured [Bosselman and Ulrich, 1984]. Optical signal transmitted through waveguides are essentially immune to electromagnetic interference, and can be configured to transmit a large number of multiplexed signals, by using wavelength, frequency, and time multiplexing methods [Tabib-Azar, 1998]. In the context of "smart microstructures" for sensors and actuators, the optical methods are potentially the most sensitive [Tabib-Azar, 1998], In addition, the rich literature in optical manipulation of signals provides a great flexibility when designing microsystems based on guided waves. Thus, Y-couplers and combiners, grating filters, microlenses, mirrors, and many interferometric systems can be produced by micropatterning optical waveguides [Tamir, 1992].

Optical signal processing and micromechanical sensing have not yet been joined productively. Several attempts to produce opto - micro-electromechanical systems (MEMS) [Jones and McKenzie, 1995] show the strong motivation to incoφorate optical means into MEMS, but present implementations are limited in functionality, or are not cost effective. The basic limitations in present opto-MEMS technology and the ways to overcame these limitations are discussed hereinunder.

Thus, an integrated optical link (IOL) on silicon, comprised of an optical source, a waveguide, and a photodetector, with all three components efficiently coupled to each other, has a tremendous impact in MEMS sensing. Moreover, the combination of such IOL with the suspended optical waveguide (SOW), as an opto-MEMS transducer is a novel approach that leads to a large number of device applications as further detailed hereinunder.

The planarity nature of micro fabrication technology and its affects on MEMS technologies: Planarity is a fundamental nature of microfabrication technology.

This property is responsible to the fact that different element in VLSI such as diodes, transistors, capacitors etc., are placed and interact in the plane of the wafer. This planarity property is inherited by the MEMS technology. and many MEMS are comprised of elements such as sensors and actuators that are fabricated and interact in the plane of the wafer and therefore have an in-plane degree of freedom (DOF). For example, the micro x-y-z stage carrying a Scanning Tunneling Microscope (STM) tip [Xu et al., 1995], the vibrating gyroscope [Maenaka et al., 1996] or the micro-gear [Legtenberg et al., 1997] are all examples of MEMS with several mechanical components interacting with each other in the plane of the wafer.

In-spite of this planarity nature MEMS with out-of-plane DOF elements are also known. The micromachined microphone made of a suspended membrane over a sealed cavity [Yazdi and Najafi, 1997], and the pendulum accelerometer fabricated using wet etch of silicon [Bergqvist and Rudolf, 1994], are examples of devices with out-of-plane DOF. However, because of the planarity nature of the microfabrication technology these MEMS have low mechanical integration abilities and therefore are capable of performing only simple tasks.

The disadvantage of the planarity nature of the microfabrication technology becomes an advantage when it comes to integrating a displacement sensor for elements with out-of plane DOF. Since these elements move out-of the plane it is relatively easy to fabricate sensors such as capacitive sensors, piezoresistive, or piezoelectric sensors on these planar elements. The movement of the micromachined microphone for example, can be sensed capacitively by coating the membrane and a counter close plane with metal, or it can be sensed by coating piezoresistive or piezoelectric materials on the membrane. On the other hand, the advantage of the planarity nature of the microfabrication technology for elements with in-plane DOF becomes a disadvantage when it comes to integration of a displacement sensor. Since elements with planar DOF move in the plane of the wafer, the traditional sensing concepts should be fabricate on their sidewalls. Not only this task is not trivial with conventional microfabrication technology, typically the overall area of the sidewall is too small to be effective. This is why comb structures, that increase the overall sidewall area, are used for capacitive sensing.

Thus, MEMS with out-of plane DOF have low mechanical integration capabilities and high sensing integration capabilities, while MEMS with in-plane DOF have high mechanical integration capabilities and low sensing integration capabilities. The disadvantages of the two technologies increase their cost and complexity for some applications. Therefore, in order for these technologies to become attractive for these applications, these disadvantages should be eliminated. To increase the mechanical integration ability of a MEMS technology with out-of-plane DOF one needs to fabricate suspended elements one on top of the other by using either multiple chip technology or using non-planar microfabrication technology. However, multiple chip technology is relatively expensive while non-planar microfabrication technology is not available yet.

On the other hand to increase the sensing integration ability of a MEMS technology with in-plane DOF one needs to associate the in-plane mechanical movement with some sensible physical property without adding fabrication complexity. The bottleneck of Opto-MEMS technology:

Optics is not common in MEMS because of the lack in the ability to, cost effectively, integrate optical components such as the light sources and photodiodes in silicon micromachining. Typically, the light source and the photodiode are introduced to the optical sensor through optical fibers. The capacitive vibrating gyroscopes mentioned above can be microfabricated using standard microelectronics and MEMS technologies. These technologies maintain the most important characteristics of microelectronics technology namely, mass production and high yield. These characteristics are not available with common Opto-MEMS technology for several limiting aspects as follows:

First, there is a lack in appropriate microfabrication technologies that allow integrative optical sensor in silicon. The reason for this is the fact that almost all of the MEMS optical sensors found today, that are used to translate mechanical effects into optical phenomena, have out-of-plane degree of freedom, while most of the MEMS technology is based on in- plane degree of freedom. Therefore in order to integrate an optical sensor with light source and photodiodes the MEMS should be based on stacking the different components one on top of the other (see for example [Degani, et al., 1998]) which is cost ineffective. Second, generally light sources cannot be fabricated in silicon. This means that light is brought to the silicon chip containing the optical sensor through an optical fiber. This task is usually done by hand and thus is very costly. Third, only limited light detectors, responding to certain wavelengths, can be fabricated in silicon. In cases where the detectors are external to the chip the same costly fiber optic integration is required. The SOW technology Suspended optical waveguides (SOW) with in-plane DOF are described in Haronian [1998a-d], which are incoφorated herein by reference]. The building block of a SOW is made of a single crystal silicon (SCS) beam with superficial layers of Siθ2:Si3N :Siθ2- The guidance properties depends also on the thickness of the different layers. For more details see Haronian [1998b]. As shown in Figures la-b, an SCS beam 30 can be fabricated with a cross section of 1.6 μm x 10 μm and may guide light with wavelength in the 1.3-1.7 μm range. The first Siθ2 layer serves as a buffer layer that allows light with wavelength in the 0.6 μm - 0.9 μm range to be guided in the Si3N4 layer. As exemplified in the Examples section that follows, the SOW was used to build several highly sensitive sensors. Figure 2a-c show light emitted form such a SOW, whereas Figure 3 shows the simplest displacement sensing application employing such a SOW. Here, a cantilevered beam-waveguide 32 is supported from one side and is connected to a proof mass 34 at its other end. The waveguide free end is tapered and is in near field proximity from a fixed tapered waveguide 36. When the SOW deforms the optical coupling between light propagating in the SOW and the fixed waveguide is modulated and is used as a measure of the SOW displacement.

Such configuration can be integrated into large MEMS such as the one shown in Figures 4-5. Here a micro-stage 38, which serves as a proof mass, is suspended by four beam elements 40 with in plane DOF. In addition, a unique interferometery based displacement sensor 41, partially integrated in micro-stage 41, was fabricated and tested [see, Haronian, 1998c, which is incoφorated herein by reference]. This planar interferometer is shown schematically in Figure 4 and in the scanning electron microscope (SEM) picture in Figure 5. Two merging waveguides 41' are integrated into a meshed frame suspended by four L shaped beams 40. In Figure 5 only two suspending beams are shown. Two fixed waveguides 42 are optically coupled into the suspended waveguides. These waveguides are fed from their other end by light that crosses the gaps between the fixed and suspended waveguides and merges at the merging section of the suspended waveguides. When the frame moves in the plane these gaps change and the phase difference of the light propagating in the two suspended waveguide is modulated. This phase modulation leads to intensity modulation at the merging section of the suspended waveguide that is recorded by a photodiode 44. Such a displacement sensor is found to be highly sensitive. Figure 6 shows the response of this sensor to shock and reveals the resonance response of the sensor. Figure 7 shows the calculated sensitivity of the sensor and compares it to the sensitivity of the simple sensor shown in Figure 3 in far field.

Due to the planar DOF of these sensors they can be tailored for various applications. For example they can be used as highly sensitive accelerometers or they can be integrated into an optical vibrating gyroscope.

It is important to note that the SOW described herein is the only one with in-plane DOF. Other SOW but with out-of plane DOF were also developed [see, Eng et al, 1995; Kim et al., 1992; Burchman et al., 1992;

Wu and Frankena, 1992: Benaissa and Nathan, 1996]. A typical example is shown in Figure 8 presenting a prior art waveguide which is made of staked layers of Siθ2-Al2θ3-Siθ2- Such waveguides are free to move out of the plane and therefore are hard to integrate with additional optical or mechanical systems. In other examples, waveguides are bounded to the membrane that is free to move out of the plane and are used to sense the displacement of the membrane. Still, as these examples show, only simple, one tasks devices, can be developed with these sensors. The integrated SOW (i-SOW)

Figure 9 schematically presents integration of optical links with fixed and suspended waveguides which is referred to herein a as integrated SOW or /-SOW. An t^'-SOW 50 according to the present invention includes an on- chip light source 54, a low-loss fixed 56 and or suspended 58 optical waveguides, an optional proof mass 60 and an on-chip photodetector 62. All components are integrated on the surface of a silicon substrate 52. For mechanical sensing applications, the waveguide is implemented with the suspended optical waveguide (SOW) technology described herein. The light source is based on GaN-based light emitting diodes epitaxially grown on silicon and on light emitting polymers for the light source.

Thus, the photodetector is based on a PIN diode that, in principle, can be fabricated in conventional silicon technology and therefore its integration is relatively simple.

An optional coupling element positioned between the light source and the waveguide and between the waveguide and the photodetector is designed according to the present invention to collect light emitted by the light source and perform "mode-conversion" so as to match the light to the waveguide mode.

An efficient /-SOW on silicone has the potential for many optical applications for mechanical sensing for silicon-based Very Large Integrated Circuits (VLSI).

Integration of solid state photodiode with silicon One of the examples of silicon based optical sensors are photodiodes. These sensors are based on a pn junction that is connected to a voltage source and to a current meter. When the pn junction is exposed to light, pairs of electron and hole are generated at the junction. These holes and electrons are swept by voltage drop on the junction and are measured by the current meter. Pn junctions are currently formed in silicon by conventional microfabrication technology using ion implantation of impurities type 'n' ('p') into the silicon substrate that is doped by 'p' ('n') type impurities. Integration of solid state light source with silicon

Major expansions in the applications of optical communication technology, silicon VLSI, sensing, and signal processing, could be made if an efficient, reliable, small, and very low cost diode laser monolithic on silicon was available. Many applications in military and civilian high rate transmission would become possible, especially those requiring large number of data paths over short distances. Thus, the search for an epitaxial material deposited on Si, that may constitute the basis for laser emission attracted a great deal of attention during the last 15 years [Fang et al., 1990]. A new possibility to build such a source involves growing III-V-Nitride compounds lattice matched to silicon [see, Salzman and Temkin, 1997, which is incoφorated herein by reference]

Material basis for solid state light sources on Si - The III-V-N System

Much effort has been expended in the synthesis and characterization of III-N compounds [Strite and Morkoc, 1992]. Alloys within this material system exhibit direct energy bandgaps ranging from 1.9 eV to 6.2 eV, forming the base for the fabrication of optical devices at wavelengths ranging from the red into the UV [Davis, 1991]. Recent progress in epitaxial growth and processing of III-N heterostructures led to the commercial production of blue LED's [Nakamura et al., 1993], the demonstration of a semiconductor laser operating in the UV [Nakamura et al., 1996], and of electronic components [Asif-Khan et al., 1994]. The III-N's exist in both the wurtzite (hexagonal) and zincblende (cubic) structures. There is at present substantial research underway on these materials and on device applications, but it is mostly directed at the hexagonal polytypes since the zincblende form of III-N binaries have only been produced successfully in a few cases. The cubic polytypes exhibit slightly lower bandgaps, and they are believed to be better suited for photonic devices due to the possibility to define optical cavities by cleaving, and the larger solubility of doping species. Alloys with more conventional III-V compounds, are an additional advantage of the cubic III-V-N system. These alloys (e.g., GaAsN, InPN, InGaAsN, etc.) are an interesting class of compounds with unusual bandgap properties and potentially important technological applications.

Growth of III-V-N materials on sapphire substrates was performed by Metal Organic Vapor Phase Epitaxy (MOVPE), in a Thomas Swan reactor that was modified for high temperature growth with NH3 precursor and their optical [Tisch et al., 1998a], structural [Tisch et al., 1998b], and electrical properties were studied and a state of the art GaN-based UV detectors were developed.

Sapphire is not an optimal substrate for III-V-N epitaxy, due to its large lattice and thermal mismatch. As a result of that, most III-V-N films produced in the world suffer from a large defect density [Tisch et al., 1998a]. However, the most striking property of GaN-based heterostructures (highly reported in the literature and also experienced in the work) is that the high defect density present in the layers does not prevent the implementation of photonic devices. This defect-immunity is in clear contrast to the high sensitivity to defects of other semiconductor families [Nakamura and G. Fasol, 1997].

LEDs were recently demonstrated with GaN-on-Si at IBM [Guha et al., 1998a]. These LEDs do not show any lifetime degradation, and they seem to be not critically sensitive to the presence of extended defects. An attractive feature of GaN-on-Si LEDs is that their electroluminescence is peaked at λ ~ 360 nm. This short wavelength allows the use of color converting dyes for the fabrication of multi-colored LEDs. The dare commercially available, and can be spun-coated on the LEDs. They are excited by the 360 nm radiation, and emit in pre-specified "down- converted" colors [Guha et al., 1998b]. Integration of solid state light source with SOW technology

Optical coupling into waveguides was studied intensively [Tamir, 1979]. Coupling elements proposed include: prisms, gratings, directional couplers, flares, etc. [Tamir, 1979]. Unfortunately, the issue was mostly studied in the context of free space to waveguide coupling, and laser to waveguide coupling [Tamir, 1979]. Having in mind that the GaN-based LED (or the Polymer Light Emitter) will be fabricated by a different technology than that of the waveguide, three specific examples of optical coupling: Butt coupling: In this case, the light emitter is an Edge Emitter

(light is coupled out through the facet of the source block and not through the surface). The waveguide is based on Siθ2, Si3θ4, technology. Optional optical configurations are depicted in Figures lOa-b for a light emitting polymer and a LED, respectively. In the configuration of Figure 10b, a coupling element which is an adiabatic mode converter fabricated using the Siθ2, Si3θ4, technology is shown. The coupling efficiency of two butt coupled elements of different geometries and refractive indices is analyzed by Hunspurger [1982, page 91, Eq. 6.2.2]

Free propagation coupling: In this case, the light source is a surface emitter located below the plane of the waveguide. The waveguide is based on Siθ2, Si3θ4, technology. Here, a filling glass with matching refractive index n_r is deposited by CVD and shaped into an elliptical shape. The center of the source and the waveguide edge are the two foci of the ellipse, so that the surface acts as a mirror that focuses the source into the waveguide edge. High reflecting coating on the elliptical surface prevents the light to escape from the coupler. n_r is given by: n_r= [n_eff( -g-)^# n_r(GaN)] l/2

Grating couplers: In this case, a periodically corrugated section on top of the waveguide cladding, or in the core-cladding interface can turn a vertically propagating beam into a horizontally guided mode, thus forming a LED-to-waveguide coupler. Also, the guided mode can be out-coupled down into an absoφtion substrate (a Si photodiode) for the waveguide-to- detector coupling. Grating couplers can be designed to produce very high coupling efficiency, leading to extremely low insertion losses in the optical link. Alternatively, a grating can be made to couple out only a small portion of the guided mode, leaving the remaining optical intensity to continue to a next detector. This is a sequential signal taper system for optical carrier distribution to many "receivers". Integration of light emitting polymer (LEP) with SOW technology:

Schematically, the structure of a simple electroluminescent device consists of a conjugated polymer layer, which acts as the emitter layer, sandwiched between two electrodes (usually, ITO on one side and Al or Ca or Mg/Ag on the other side), having different work functions. Upon application of a voltage across the cell, electrons are injected into the conduction band (LUMO or π * band) of the polymer in the vicinity of one electrode and holes are injected into the valence band (HOMO or π -band) in the vicinity of the other electrode. The oppositely charged carriers may then, if they meet, form an exciton and recombine with the emission of light. Although here electrons and holes are referred, it will be appreciated that in conjugated polymers other charged excitations such as those coupled to lattice distortions and referred to as positive and negative polarons are likely to participate. The emission properties of such devices strongly depend on the mismatch of the energy barriers at the polymer/electrode interfaces, the mobility of the polymers and its homogeneity.

Devices such as described above where realized first by the group of R. Friend and D.D.C. Bradley in Cambridge [Burroughs et al., 1990]. Today, some of the PPV-based LED devices have demonstrated external efficiencies of ~ 3-5 % (higher than 15 Lumens/Watt) and lifetimes exceeding 4000 hours under DC operation. This is achieved by modification of the interfaces (using an ultrathin oxide layer or organic layers), using additional electron and hole transport layers and encapsulation. The potential advantage of using PPV and/or derivatives as opposed to inorganic compounds lies in the large number of colors available by chemical control and the possibility of producing large and flexible displays as a result of the mechanical properties of the polymers. Recent degradation studies have indicated that the main mechanisms for degradation of such LEDs are photo-oxidation and heating of the device [Sawate'ev et al., 1997]. The former can be dramatically improved by encapsulation while the later can be avoided by applying short electric pulses instead of DC operation. The area of polymer-based LED can be changed from large area of, for example, 100 cm^ to nano-size pixels. Microfabrication of array of LED pixels, each of 20 micron size have been demonstrated [Noach et al., 1996]. Recently several groups have demonstrated lasing from polymers under optical pumping using intense picosecond laser pulses [Tessler et al., 1996]. These were achieved using a microcavity or a waveguide arrangement. In the former case, the emitting layers were sandwiched between two flat mirrors such as a Distributed Bragg Reflector (DBR) or a metallic films. The observed narrowing of the emission spectral lines and the improved directionality of the emitted light were taken as evidence for lasing, super-radiance or superfluorescence due to collective excitations. Clearly, these very important experiments indicate that conjugated polymers can be considered as highly emissive solid state materials and as laser media. It remains technologically much important to achieve electrically- stimulated lasing. Technically, this may be done by inserting a transparent conducting layer (such as ITO) on top of the reflector. However, a prerequisite for any progress in this direction is the observation of large transient currents under very short electrical pulses. Estimates based on the present knowledge indicate that a current density of the order of lO^-lO^ A/cm^ may be needed to reach the excitation threshold for lasing. Recent time-resolved EL experiment [Chayet et al., 1997; Tessler et al., 1998] using a simple EL device composed of a single layer PPV sandwiched between electrodes have demonstrated that the application of short and strong electrical pulses (100 nsec to 10 nsec, amplitude of up to 1 kV) leads to extremely strong optical burst. The maximum current densities during the EL emission is 10^ A/cm^ and peak brightness of lO^-lO^ Cd m^ can be achieved easily.

Today, almost all of the research on organic-based LED devices is concerned with injection LEDs described above. In this type of devices carriers are injected from the electrodes, move under the action of the biased electric field, form excitons which may recombine radiatively. Such devices requires thin films, the optimal thickness is 100 nm, to yield low threshold voltage and enough brightness.

Miniaturization and possible electrically pumped lasers on silicon substrate:

Several methods are adapted to fabricate micron-sized LEDs. One is to adapt silicon chip fabrication technology. Such a technology and chips are available today and are extensively being used for liquid crystal display devices. Pixels of several microns in diameter are easy to achieve using this technology. The LEP will be shaped into these micrometer sized pixels. Another method is to use lithographic methods and laser ablation technique as described in [46]. This method allows the fabrication of array of micron- sized pixels.

There is an impressive progress in the development of optically pumped lasing from conjugated polymers using different configurations such as waveguides [Diaz-Garsia et al., 1997] "whispering mode" microrings [Nabor et al., 1998] microdisks and polymer-coated wires [Frolov et al., 1998]. Needless to mention, one or several of the above configurations may be used for electrically pumped laser by incoφorating ITO transparent electrode. To achieve lasing by pumping electrically, high peak current density and consequently very strong electric pulses are necessary. These may lead to a device degradation. It is absolutely necessary, therefore, to work under very short (sub-nanosecond) and strong (10 MV/cm) pulses to reduce degradation effects. Integration of LEDs with the SOW technology:

One of the issues that are important in the integration of a light source, e.g., LED, and a SOW is the coupling between the light source and the SOW. First one notes that the index of refraction of Si3N4 at 600 nm is about 2. In addition one notes that it is possible to fabricate LED with index of refraction of 2. Therefore if the LED is placed in contact with the Si3N4 waveguide (butt coupling) one can expect an efficient light leakage form the LED into the Si3N4 waveguide. In Figure 11 the LED is integrated at the end of the Si3N4 waveguide. In order to avoid oxidation, the LED is coated with thin (0J μm) layer of Siθ2 with index of refraction of 1.5. This layer separates the LED form the surrounding with index of refraction of 1. Since this isolating layer is rather thin one can assume that the effective index of refraction of the medium surrounding the LED is actually smaller than 1.5. Therefore one can assumption that only small amount of light will leak form the edges of the LED, and therefore by shaping the LED with an appropriate tapering shape it is possible to channel the light into the Si3N4 waveguide. Since the LED is based on polymers such shaping is possible using, for example, oxygen plasma.

EXAMPLE 2 In-plane degree of freedom (DOF) optical waveguide displacement sensors based on geometrical modulation

Suspended optical waveguide (SOW) displacement sensors technology is presented. The sensors are based on optical modulation in the form of energy losses and mode conversions, resulting from relative displacement of aligned and suspended waveguides. The building block of the suspended waveguides is a single crystal silicon (SCS) beam with on top layers comprising of a 0.6 μm thick Siθ2, 0.4 μm thick Si3N4, and 0.6 μm thick Siθ2- The SCS has a typical cross section of 1.6 μm x 10 μm and may guide light with wavelength in the 1.3-1.5 μm range. The first Siθ2 layer serves as a buffer layer that allows light with wavelength in the 0.6 μm -0.9 μm range to be guided in the Si3N layer. This Example discusses the theoretical consideration, the fabrication process, and the characterizations of fixed-free and fixed-fixed SOW configurations.

Displacement sensing using suspended waveguides is not a new concept [Eng et al., 1995; Kim et al., 1992; Burchman et al., 1992; Wu and Frankena, 1992; Benaissa and Nathan, 1996]. Suspended waveguide sensing is based on physical phenomena that can be grouped into several methods: Geometrical Modulation (GM), Evanescent Field Modulation (EFM), and Index of Refraction Modulation (IRM). Eng et al. [1995], fabricated suspended silicon beams for light guiding in the 1.3 μm range. Kim et al. [1992], used a suspended dielectric film hanging inside the evanescent field of a waveguide in order to modulated the light passing through the waveguide (EFM method). Burcham et al. [1992] used a silicon nitride waveguide over a suspended silicon structure to modulate the light coupling between the suspended waveguide and a fixed waveguide aligned to the silicon structure (GM method). S. Wu, and H. J. Frankena [1992] fabricated a fixed-fixed waveguide and a cantilever waveguide made of an AI2O3 layer sandwiched by two S1O2 layers for displacement sensing using the GM method. K. E. Burcham, et al. [1992], used the GM method to fabricate an accelerometer. K. Benaissa and A. Nathan [1996] fabricated waveguides in a Mach-Zehnder configuration on top of a silicon membrane for optical sensing of the membrane vibrations. In this case the sensing of the membrane vibration is mainly due to the stress induced index of refraction modulation (IRM method).

In all of the above methods the waveguides have an out-of-plane degree of freedom. Due to the planar nature of current microfabrication technology, the restriction to vertical displacement limits the design flexibility and the integration with other MEMS. Such out-of-plane technology requires vertical stacking of optical elements such as waveguides, sensors that are hard to implement.

The SOW technology, presented herein, is based on waveguides with in-plane degree of freedom which allows an in-plane interaction and integration with MEMS. The geometrical modulation of fixed-free and fixed-fixed SOW are discussed. Sensors, in fixed-free and fixed-fixed configurations, are tested and their integration with micromachined actuator is demonstrated. Optical modes in suspended waveguides:

The cross section of the building block for the SOW displacement sensor is shown schematically in Figure 12. The cross section comprises of a single crystal silicon (SCS) coated with 0.6 μm thick Siθ2, 0.4 μm thick S13N4, and 0.6 μm thick Siθ2- The width of the beam is typically in the range of 1-2 μm, and the height is in the range between 8-14 μm. This building block contains two possible guiding paths. First, the silicon itself with index of refraction around 3.5 at wavelength of 1.3 μm, may guide light in the 1.3-1.6 μm range [Soref and Lorenzo, 1986]. In addition, the Si3N4 with index of refraction around 2 at wavelength of 0.6 μm, may guide light in the 0.6-0.9 μm range [Stutius and Streifer, 1977]. This guidance is possible because the Siθ2 layer underneath blocks the light guided in the Si3N4 from leaking into the silicon. In general, the Siθ2 layer should be relatively thick. Nevertheless a thick Siθ2 layer complicates the fabrication process. As the evanescent field of light propagating in the Si3N4 layer extends about 0.5 μm into the Siθ2 layer it is assumed that 0.6 μm of Siθ2 is sufficient for waveguides that are typically few hundred microns long.

Rectangular waveguides were analyzed by E. A. J. Marcatili [1969]. Here the Marcatili's analysis is followed to find the guidance properties of the silicon and in the Si3N4 waveguides as a function of their dimensions. Figure 13 shows the waveguide-beam cross section. The analysis for the TE and TM modes is similar, so TE modes propagating along the waveguide is assumed. The solution to Maxwell equations for the silicon waveguide yields the following transcendental equations for the transverse propagation constants: k_J,^pa = pπ - tan-¹ k_x ^pξ₃ - tan-^I k_ις ^pξ₅ (4) k ^qb = qπ - tan^"' - tan^"1 (5)

where p is the mode in the 'y' direction and q is the mode in the direction, k ^p = k ' x l ^p = k x2 ^p = k, x4 ^"

' ⁼Tⁿ" λ is the wavelength in vacuum, n_j and k_j are the index of refraction and the propagation constant in region 'i' respectively, and η_j, ξ_j are the penetration depth of the field in region .

The propagation constant of the pq mode is therefore:

Assuming

with i=3,5 for the silicon waveguide k. ^qA. β,^q = « 1 (8) π with i=2,4 for the silicon waveguide. λ where, A, = - , 2 2 \'/2 i-1-5

!(n, - n, )

P - P -

The solutions for k_χ , k_y and thus for k_z are in the form:

A₃ +A₅

= ^(l + (9) a v πa J

Considering the silicon waveguide configuration as shown in Figure 13, one defines the guidance parameter:

G„ = k.ⁿ ) k,² - k₅ ²

The guidance parameter measures to what extent a given mode pq pq described by k_z propagates along the waveguide. If k_z =k$ the parameter is zero. This means that the waveguide guides the light to the same extent as the surrounding medium and thus the light in the waveguide pq will radiate. If, on the other hand, k_z =k_j>k5, the guidance parameter equals 1 , and the mode is strongly guided.

Figure 14 shows G15 as a function of the width of the silicon waveguide, a, for different modes. The analysis is carried out for b=10 μm and λ=1.3 μm. It is important to view this Figure together with Figure 15 that shows parameters α_j , and β_j , as function of the width. One notes that these parameters do not meet the assumptions in Equations 7 and 8 for every waveguide width. This results in a deviation from the graphs that are shown in Figure 14. Since one is interested only in cases where light is guided in the waveguide, one restrict the discussion to the case where G is high enough. Marcatili choose G=0.5 as a limiting value. For G>0.5 one finds from Fig. 15 that the conditions in Equations 7 and 8 are meet. As is revealed from Figure 14, for cross section of 1.6 μm x 10 μm the silicon beam is a multimode waveguide in the 1.3 μm range. Figure 16 describes G12 as a function of the Si3N4 waveguide height, b, for different modes. The analysis was carried out for a=1.6 μm and λ=0.633 μm. Note that unlike the silicon waveguide analyzed above, the different modes do not overlap. From Figure 16 one observe that for b=0.4 μm there is one mode in the x direction and three modes in the y direction. It is possible to show that Equations 7 and 8 hold for these modes.

In is important to note that the modes in the waveguides above are more confined than in a typical waveguide fabricated inside or on top of a substrate. This is because of the cladding of the letter that increase the cut- off conditions. In general it is possible to clad the suspended waveguide with appropriate material, such as silica, and in such a way to control the guiding properties of the waveguide.

Geometrical modulation (GM) in waveguides:

As shown above the suspended waveguides may guide several modes. In the analysis below one assume an efficient coupling between the input fiber and the sending waveguide, such that practically all the energy is guided by the first mode. In that case the displacement δ of the suspended waveguide is sensed through the modulation of the coupling between the first mode in the sending waveguide and modes in the receiving waveguide. One denotes.the mode coupling by Cι _u(δ), and Cι_v(δ), where the subscript 1 refers to the first mode of the sending waveguide, and the subscript u,v refer to the symmetric and asymmetric modes, respectively, in the receiving waveguide. As is shown below, with 2 μm gap between the two waveguides, and with δ=0, the first modes, Ci _j (6=0), are highly coupled. There is a non zero coupling between the first mode and the symmetrical modes (e.g.,

while the coupling between the first mode and the symmetrical modes is practically zero(e.g.,

C 14(6=0)). Therefore one can state that 1-Cι ι (δ) represents the energy losses as a function of δ, by the first mode in the receiving waveguide and C_{j u}(δ) with u>l, and Cι _v(δ), represent energy conversion from the first mode in the sending waveguide to higher modes in the receiving waveguide. As is further shown below this higher mode interaction is relatively small compared with the energy lost. Nevertheless, this mode conversion may lead to speckle effects that, depending on the sensing configuration, may contribute to the signal.

As shown in Figures 17a-b, displacement in the y direction induces the interaction of the modes in the 'y' direction. The symmetric and asymmetric modes in the 'y' direction can be respectively described as: E_u(y) = A_uCos(k_yily) (12)

E_v(y) = A_vSin(k_yvy) (13)

The u,v notations are equivalent to the p notation in section 2. The first mode emerging from the sending waveguide is subject to a Gaussian expansion that can be described by:

E_s(y) = A_sExp[-y² /w,²] (14)

where w, (z₀) = w ΌI 1 and WQ I is the Gaussian width at Z=0

such that the coupling between the first guided mode and the propagating beam,

fCos(k_lyy)Exp[- y ² /w₀₁ ² ]dy

{Cos² (k_lyy)dy }Exp[-2 y7w_0l ² ]dy

is maximized.

The modes propagating in the receiving waveguide correspond to

Hermit-Gaussian beam modes propagating in the surrounding medium and can be described by: E_gu,v(y) = H_u,NV2y/w_U J Exp[- y²/w_{u v} ² ] (15)

where, H_{u v} is Hermit polynomial, _{u v} ² (z₀ ) and

WQ_U and WQ_V respectively satisfy:

/Cos² (k _uxx)dy JExp[-2 y ² /w_0u ² ]dy

(16)

In that way one can introduce the coupling between the first mode in the sending waveguide and the modes in the receiving waveguide as the coupling between the first mode after going through a Gaussian expansion at Z=ZQ and the Hermit-Gaussian beam modes at Z=0. When the suspended waveguide is subject to a displacement, δ, this coupling can be described as:

J Exp[-(y - δ)² /w ,² ]H_{U v} (V2y/w_0u , )Exp[- y ² ' w 0u,v ]dy|

C_lB..(δ)

|Exp[-(y - 6)²/w,² ]dy jH_U]V ² (V2y/w_{0u v})Exp[-2y²/w_{0u v} ²]dy

(17)

Figure 18 is the coupling Cι _{u v} as a function of the displacement 6, for a silicon waveguide with the following parameters: a=1.6 μm, b=10 μm, zø=2 μm, and λ=1.3 μm. As shown in this Figure most of the interaction is carried by the first modes (C_j ι ). In addition, a constant increase in the displacement reduces the coupling of the first mode with the symmetrical modes, while leading to a peak in the coupling with the asymmetrical modes. This is explained by the fact that the energy of the asymmetrical mode is conserved mostly off axis. Figures 19 and 20 are the sensitivity, dC_{}u v}/d6, and the relative sensitivity, dCι _{u V}/Cι _{u v}dδ respectively, for the same parameters as above. As shown in Figure 19 the sensitivity of the first modes, dC_{j j}/dδ, peaks when the displacement is half of the waveguide width. A sensitive configuration will therefore result when the sending waveguide is initially displaced by half of the waveguide width. The relative sensitivity of the coupling with the first mode is linearly with the displacement, but deviates from linearity for the higher symmetrical modes (i.e., CR-13, CR-15). cR-12,

on the other hand overlap at small displacement, peak at about δ=0J μm, and decreases as the displacement increases. At large displacement CR-12, and

split and merge with the relative sensitivities of the symmetric modes.

The geometrical modulation analysis described above assumes the waveguides width is constant. Nevertheless, it is possible to consider the interaction between the two waveguides where at least one of them has a tapered end, as shown in Figures 21a-b. This kind of interaction can be very sensitive if, like in scanning tunneling optical microscope (STOM), the tips are shaφ enough and the interaction takes place in the near field region. This type of interaction is another flexibility an in plane configuration has over an out of plane configuration since it is not trivial with current microfabrication technology to form vertical tips. The exact interaction of tip and sample in STOM configuration can not be solved analytically. Fabrication process:

SOW displacement sensor is fabricated using the SCREAM [Shaw et al., 1994] (single crystal reactive ion etching and metalization) process. Figures 22a-h depict a step by step flow diagram of the fabrication process. The process starts with a deposition of Siθ2(0.6 μm), Si3N4(0.4 μm), and Siθ2(2 μm). The device image is transferred to the wafer top using optical lithography, and the three layers are etched using RIE with CHF3 (22b-c). The image is further transferred into the silicon bulk using RIE with Cl2:BCl3 (22d). The depth of this etching step is typically in the range of 8- 10 μm. Next, the pattern is passivated using PECVD of S.O2 (22e), and the floor oxide is etched using RIE of CHF3 (22f). The pattern is etched again by RIE with CF^BC^ chemistry (22g). This step exposes bare silicon under the passivated silicon, that is etched by an isotropic RIE with SF^ (22h). This results with structures suspended 3-4 μm from the substrate. The top 2 μm Siθ2 is used as a mask for the process. Typically about 0.6 μm of this layer remains at the end of the process.

The quality of the sidewalls of the silicon waveguide formed during the fabrication can induce losses of guided modes to radiating modes. The surface roughness is estimated to be below OJ μm, leading to losses smaller than 0.2 db/mm.

Testing setup:

The sensors are excited acoustically and mechanically using a mechanical shaker. Figures 23a-b shows schematically the testing setup of the mechanical excitation. The shaker is connected to the round base of the testing apparatus that is hung by three rubber springs. In order to excited the devices at different angles the testing apparatus base may be rotated and connected to the shaker at different angles relative to the tested device. A calibration accelerometer with a flat response in the 0-70 kHz range is connected to the shaker arm. The spherical lens at the end of the input fiber, the output fiber, and the sample are mutually manipulated using three 5 axis manipulators. The output fiber feeds a low noise photodiode, which in turn feeds a spectrum analyzer. Device configuration and testing results:

Figure 24 is a scanning electron micrograph showing the general view of a chip containing several sensors in a fixed-fixed configuration. The sensors are fabricated on top of a plateau formed by KOH etch. The distance between the input/output ports of the waveguides and the edge of the plateau is typically between 20 μm and 100 μm. The plateau is about 100 μm above the lower plane, which allows optical fibers with diameter smaller than 200 μm to approach the input/output ports of the waveguides.

Figure 25 is a scanning electron micrograph of a fixed-free SOW displacement sensor with a proof mass at its end (An electrode is fabricated close to the proof mass for applications that combines electrostatic forces). As shown in the close view in Figure 26 the waveguide is suspended from a parabolic horn structure that expends from 1.6 μm to 10 μm. It is found that such parabolic transition between waveguides can be designed to operate adiabaticaly, where the lowest mode propagates through the parabolic section without transferring power to higher modes [Milton and Burns, 1997]. Therefore, in addition to supporting the waveguide, the parabolic horn couples light efficiently into and out from the waveguide. Also shown in Figure 26 are two blocking ribs placed vertically to the waveguide, that block scattered light from reaching the output fiber. In order to increase the sensitivity of the sensor a tip end is used, as shown in the close view in Figure 27. The radius of the tips is estimated to be about 0J μm. As was described above, the cut off conditions are not meet below certain waveguide width and therefore it expected that part of the light will radiate through the tip walls before reaching the tip.

Figure 28 is the response of the device to acoustical excitation. The upper curve in this Figure is the vibrational frequency response when the device is positioned below the optical path. This spectrum contains a noisy DC signal resulting from the laser light collected by the photodiode. As the sensor, aligned to the light path, is raised up and blocks the light path the DC signal drops and a resonance signal appears as shown in the lower curve. The resonance frequency of 3.4 kHz approximately equals the calculated resonance. Figure 29 is the response of the same device to mechanical white noise vibration. Three harmonic peaks appear in the lower section of the spectrum. A similar response was observed when one of the suspending rubber spring was forced to vibrated. This effect may indicate that the peaks observed are the response of the sensor to the vibrational modes of the suspending springs.

It is important to note that in the case of the sensor above the expected frequency response should be twice the resonance frequency as in one vibrating cycle the amplitude at the output port changes from maximum to minimum twice. This response was not observed in the experiment. The reason for that is the fact that an initial misalignment between the sending waveguide and the receiving waveguide was observed (see close view in Figure 27). Since the vibration amplitude in the experiments above is smaller than this initial misalignment the signal at the output port follows the mechanical excitation, and thus the frequency response equals the resonance frequency.

Figure 30 is a scanning electron micrograph of a fixed- fixed version of the geometrical modulation sensor. The structure is made of a fixed- fixed beam with a proof mass held by the tips structure at the center. As a result of vibration the two tips are angularly displaced relative to each other leading to an optical modulation. The bending and torsion resonance frequencies of this structure are estimated to be 6.2 kHz, and 39 kHz, respectively. Figure 31 is the response of the sensor to acoustical excitation, revealing the bending resonance.

Thus, this Example presents a detailed analysis and test of an in- plane degree of freedom SOW displacement sensor based on Geometrical Modulation (GM). The theoretical analysis showed that from the geometrical point of view properly isolated dielectric layers or beams with typical dimensions used in MEMS are multimode waveguides. The sensing concept was demonstrated by fabricating and testing fixed-free and fixed- fixed SOW displacement sensor. The in-plane degree of freedom, as oppose to out of plane degree of freedom, increases the fabrication flexibility: The sensor described in Figure 30, for example, is not trivial to implement using out of plane configuration.

Dielectric layers and beams can be used as optical waveguides and can be tailored into MEMS where they can be used as an independent displacement sensor. Figure 32 is a scanning electron micrograph showing a micro platform suspended over four bending beam. The suspending beams serve also as suspended waveguide with tip-tip configuration as shown in the close view in Figure 33. The platform can, for example, manipulate a tip or a mirror or some other element, and the waveguides on its sides can be used to independently sense its displacement.

EXAMPLE 3

Displacement sensing using geometrical modulation in reflection mode

(GM-RM) of coupled optical waveguides

Another optical displacement sensing method for micro-electro- mechanical systems (MEMS) is presented in this Example. The concept presented herein uses the displacement of micromachined structures to modulate the optical coupling between two waveguides fabricated next to the structure. The waveguides are based on suspended optical waveguide displacement sensors technology and are fabricated in parallel with the suspended structure using the single crystal reactive ion etching and metalization (SCREAM) process. The building block of the waveguides is a single crystal silicon (SCS) beam with superficial top layers comprising a 0.6 μm thick Si0₂, 0.4 μm thick Si₃N₄, and 0.6 μm thick Si0₂- The SCS beam is fabricated with a cross section of 1.6 μm x 10 μm and may guide light with wavelength in the 1.3-1.5 μm range. The first Siθ2 layer serves as a buffer layer that allows light with wavelength in the 0.6 μm -0.9 μm range to be guided in the Si3N4 layer. This Example discusses the response of the sensor to acoustical pulse excitation.

The geometrical modulation in reflection mode (GM-RM) concept is motivated by two optical sensing concepts: The reflection near-field microscopes [Cline and Isaacson, 1995; Jalocha, 1993] and the fiber optic microphone [Garthe, 1993; Paternottre et al., 1993]. In the reflection near- field microscope light emitted from a tip located inside the near field region from a sample is reflected from the sample into a receiving optical detector. As the sample is scanned, the change in the topography modulates the intensity of the reflected light reaching the optical detector. This modulation is used to reconstruct the surface topography. Using similar concept but in far field, the fiber optic microphone uses light reflected from the microphone membrane to sense the exciting acoustical field.

The GM-RM concept is implemented by fabricating two fixed waveguides next to a suspended structure having planer degree of freedom. The two waveguides are fabricated at an angle to each other such that light emitted by one waveguide is reflected by the suspended structure and is coupled into the second waveguide. As the structure moves, the coupling efficiency between the two waveguides is modulated, and is used as a measure to the suspended structure displacement. As the input and output waveguides are located close to the device (see below) the local displacement of the structure is sensed. Device configuration:

The cross section of the building block for the SOW displacement sensor is shown schematically in Figure 34. The cross section comprises a single crystal silicon (SCS) coated with 0.6 μm thick S-O2, 0.4 μm thick Si3N4, and 0.6 μm thick Siθ2- The height of the beam is in the range between 8-14 μm. The width of the beam depends on the specific puφose of the design. Suspended waveguides [Haronian 1998b-c] have cross section in the range of 1-2 μm. Fixed waveguides, such as the waveguides to be used in the GM-RM described herein, are typically 10 μm wide, and are tapered at their end. The building block depicted in Figure 34 contains two possible guiding paths. First, the silicon itself with index of refraction around 3.5 at wavelength of 1.3 μm, may guide light in the 1.3-1.6 μm range [Soref et al., 1986]. In addition, the Si3N4 with index of refraction around 2 at wavelength of 0.6 μm, may guide light in the 0.6-0.9 μm range [Stutius and Streifer, 1977]. This guidance is possible because the Siθ2 layer underneath blocks the light guided in the Si3N4 from leaking into the silicon. In general, the Siθ2 layer should be relatively thick. Nevertheless a thick Siθ2 layer complicates the fabrication process. As the evanescent field of light propagating in the Si3N4 layer extends about 0.5 μm into the Siθ2 layer one assumes that 0.6 μm of Siθ2 is sufficient for waveguides that are typically few hundred microns long. This guidance was demonstrated elsewhere [Haronian 1998a]. To test the concept of the GM-RM a meshed plate suspended by four beams is fabricated as shown in Figure 35. The beams have a cross section of 1.6 μm x 10 μm and are L shaped in order to relief internal stress. Next to the meshed frame are two tapered waveguides. The tips of the transmitting and receiving waveguides are located 1 μm, and 6 μm away from the suspended structure, respectively (see Figure 36). The radius of the tips is estimated to be 50 nm.

SOW displacement sensor technology is realized by using SCREAM [Shaw et al., 1994] (single crystal reactive ion etching and metalization) process, that combines an-isotropic and isotropic dry etch to define and release silicon based structures. A description of the SOW displacement sensor technology is given in reference [Haronian 1998b]. As shown in Figures 35 and 37a-b, the sensor is fabricated on top of a plateau formed by KOH etch. The distance between the input/output ports of the waveguides and the edge of the plateau is about 20 μm. The plateau is about 100 μm above the lower plane, which allows optical fibers with diameter smaller than 200 μm to approach the input/output ports of the waveguides.

The interaction between the two tips does not have an analytical solution. A finite element analysis (FEA) using a non conventional method was used to analyze tapered waveguides interaction in in-axis configuration. This analysis showed sensitivity in the order of 10 μW/μm. If one assumes the NEP of the sensor is about 10"9 W, then displacements as small as 1 Angstrom can be measured.

Experiments and results: The GM-RM sensor is excited acoustically and is tested using light with wavelength of 0.633 μm that is guided by the Si3N4 layer, and using light with wavelength of 1.3 μm that is guided by the silicon beam. Figure 38 is a schematic view of the testing setup. The input and output fibers have a 9 μm and a 50 μm core diameter, respectively. The input fiber is coupled into the waveguide through a spherical lens and the output fiber is located next to the output port of the waveguide. The fibers and the sample are mutually manipulated using 5 degrees of freedom positioners. The output fiber is connected to a low noise photodiode, which in turns is connected to a spectrum analyzer and to an oscilloscope. The input fiber is chosen to have a 9 μm core diameter for both light sources. Therefore, in order to switch from one light source to the other only the light source and the photodiode are replaced. This leaves the input and output fibers aligned to the sensor. The sensor was excited acoustically at low frequency by applying a square waveform to the loudspeaker at frequencies ranging between 1 Hz and 100 Hz (see Figure 38). As the resonance frequency of the sensor is much higher than frequencies in this rage a pulse response was observed. The sensor response to pulse excitation can be described as [Thomson, 1971]: x(t) = ∑ C, exp[-ωJ/Q₁ ]cos( /l - l/Q² ω J + φ_i ) (18) where, 'i' is index describing the vibration mode C_j is a constant, α>_j is the natural frequency, Q_j is the quality factor, and φ_j is the phase. Figure 39 is the time response of the sensor to acoustical pulse, and

Figure 40 is the response as recorded by a spectrum analyzer. The response contains the main resonance frequency at f_r=3.7 kHz, and two signals at 2f_p and at 3f_r. f_r and 2f_r can be easily observed in the pulse response shown in Figure 6. The device was designed to have one in-plane mode at 3.8 kHz. This mode is identified as f_r in Figure 40. In addition, as a by product of the SCREAM process, the device is stiff to out-of-plane displacements. Nevertheless, due to the beams L shape, and their connection to the meshed plate several additional modes are possible. Figure 41 shows the first 4 vibration modes of the suspended structure calculated using FEA. As was expected, and as this analysis shows that higher modes are not harmonics of the first translation mode. It is therefore assumed that the signals detected at 2f_r and at 3f_r are due to nonlinear response of the device, or due to the non symmetrical time response, around the time axis as shown in Figure 39. Thus, in this Example a displacement sensor is described that uses the geometrical modulation in reflection mode (GM-RM) of coupled waveguides. The sensor is based on measuring the light reflected from a free to vibrate structure. The light is guided, emitted, and guided again by two waveguides fabricated next to the structure, which allows local displacement sensing. The device was tested by pulsed acoustical excitation and its resonance frequencies were measured and compared to a finite element model.

The RM-GM sensor is similar in it concept to the fiber optics microphone. On the other hand, two major elements distinguish it from the reflection near filed optical microscope. First, in the near field microscope the tip and sample interact in the near field which is in the 500 nm range in case of 1.3 μm light. In addition, the tapered part of the tip is coated by metal in order to force light emission only from the tip. This creates a highly localized light source which is important for the performances of near field microscopes. In the RM-GM sensor the tip of the sending waveguide is located about 1 μm away from the vibrating element, and currently the taper part of the waveguide is not coated. While the metallic coating is required for high localization of the light source, the near field proximity is necessary if ultra high sensitive displacement sensor is required. Metallic coating can be integrated into the fabrication process. On the other hand, it is not practical to fabricate the tip and the vibrating device in near field proximity without addition of active adjustment. Practically, it is possible to fabricate the RM-GM sensor such that the position of the vibrating device or the input waveguide can be adjusted electrostatically.

Although the GM-RM sensor suffers from low integration, due to lack of light sources in silicon, it has several advantages over the more conventional sensing means. Unlike piezoelectric or piezoresistive sensing, and as in many optical sensors, the GM-RM sensing concept does not requires physical interacts between the sensing elements (i.e., the waveguides) with the sensor itself (i.e., the suspended structure), and thus does not affect its mechanical properties. In addition, the GM-RM does not require a second electrode such as in capacitive sensing, and therefore is not exposed to problems such as sticking and squeeze film effects.

EXAMPLE 4 Geometrical modulation based interferometry for displacement sensing using optically coupled suspended waveguides

A geometrical modulation based interferometry (GMI) for displacement sensor is presented in this Example. The implementation of the GMI is based on the suspended optical waveguide displacement sensors technology. The interferometry effect of the GMI results from light propagating in geometrically modulated, and mutually coupled suspended waveguides with in-plane degree of freedom. The building block of the suspended waveguides is a single crystal silicon (SCS) beam with superficial layers comprising a 0.6 μm thick Siθ2, 0.4 μm thick SJ3N4, and 0.6 μm thick Siθ2- The SCS beam is fabricated with a cross section of 1.6 μm x 10 μm and may guide light with wavelength in the 1.3-1.5 μm range. The first Siθ2 layer serves as a buffer layer that allows light with wavelength in the 0.6 μm -0.9 μm range to be guided in the Si3N4 layer. This Example discusses the theoretical consideration, and the characterizations of a GMI displacement sensor.

A general discussion about displacement sensing using suspended waveguides is given in reference [Haronian 1998b]. Waveguide displacement sensors are based on several physical phenomena that can be divided into several groups: geometrical modulation (GM), evanescent field modulation (EFM), and index of refraction modulation (IRM). GM is a general name for direct modulation of optical properties as a result of the mechanical movement of the sensor. For example, optical modulation resulting from vibrating mirrors such as the 'mirrors on chip' of Texas Instruments, or optical modulation in the form of energy losses and mode conversions resulting from relative displacement of aligned and suspended waveguides. Interferometry displacement sensing is typically based on Mach-Zehnder Interferometer (MZI) or on Fabry-Perot Interferometer (FPI). The MZI is used such mat one arm of the interferometer is subject to mechanical excitation leading to a phase difference between light propagating along the interferometer arms. For example, Benaissa and A. Nathan [Benaissa and Nathan 1996] fabricated waveguides in a Mach- Zehnder configuration on top of a silicon membrane to optically sense the membrane vibrations. The sensing in this case is mainly due to the stress induced index of refraction modulation (IRM configuration). The FPI is based on the modulation of the interferometer cavity length. For example, A. Tran et al., used the FPI to realize optical accelerometers and optical filters [Tran et al., 1995, 1996]. Both, the MZI and the FPI displacement sensors, are based on out- of-plane displacement of the active element in the interferometer. In this paper we discuss a geometrical modulation based interferometry (GMI) displacement sensor that is different from the two, above mentioned, interferometry mechanisms. The sensor is based on SOW displacement sensor technology [Haronian 1998b; Soref and Lorenzo, 1986] with waveguides having an in-plane degree of freedom. One major advantage of the SOW displacement sensor technology is the in-plane degree of freedom given to the suspended waveguides. The in-plane degree of freedom allows the waveguides to interact with each other and with MEMS. For example the interferometry concepts discussed in this paper cannot be implemented with suspended waveguides having an out-of-plane degree of freedom, as it is not practical to fabricate two suspended waveguides one on top of the other with current microfabrication technology. The GMI displacement sensor can be used for acoustical, flow and mechanical displacement sensing. This Example discusses the theoretical considerations, the fabrication process, and the response of the sensor to acoustical excitation. General considerations of the SCS:

building block:

The cross section of the building block for the SOW displacement sensor is shown schematically in Figure 34. The cross section comprises a Single Crystal Silicon (SCS) beam coated with 0.6 μm thick Siθ2, 0.4 μm thick Si3N4, and 0.6 μm thick Siθ2- The width of the beam is typically in the range of 1-2 μm, and the height is in the range of between 8-14 μm. This building block contains two possible guiding paths. First, the silicon itself with index of refraction around 3.5 at wavelength of 1.3 μm, may guide light in the 1.3-1.6 μm range [Soref and Lorenzo, 1986] . In addition, the Si3N4 with index of refraction around 2 at wavelength of 0.6 μm, may guide light in the 0.6-0.9 μm range Stutius and Streifer, 1977 ; Haronian 1998a. This guidance is possible because the Siθ2 layer underneath blocks the light guided in the Si N4 from leaking into the silicon. In general, the Siθ2 layer should be relatively thick. Nevertheless a thick Siθ2 layer complicates the fabrication process. As the evanescent field of light propagating in the Si3N4 layer extends about 0.5 μm into the Siθ2 layer we assume that 0.6 μm of Siθ2 is sufficient for waveguides that are typically a few hundred micrometer long.

Rectangular waveguides were analyzed by E. A. J. Marcatilϊ [1969], and Marcatili's analysis was applied to the SOW displacement sensor technology [Haronian 1998b]. It is found that the 1.6 μm x 10 μm SCS beam in the building block described above, is a multimode waveguide for light with wavelength of 1.3 μm, and the 1.6 μm x 0.4 μm Si3N4 layer is a single mode waveguide for the transverse direction, for light with wavelength of 0.6 μm. The GMI displacement sensor:

The general concept of the GMI is described schematically in Figure 42. Light is divided between two fixed waveguides aligned to suspended waveguides at distances z_a, and z^. As a result of a displacement δ of the supporting frame the gaps z_a, z_j- are modulated, and the phase gained by the light waves after crossing the free space is:

where nn. is the index of refraction in the surrounding medium. As light emerges form the waveguide into the surroundings it goes through a Gaussian expansion, WQ is the width of the Gaussian at the waveguide- surrounding medium interface.

The amplitude and intensity of the light at the merged waveguide respectively are:

A₀ = A_0a exp[i(<yt + φ_& )] + A_0b exp[i(<* + φ_b )] (20) 0 ^— 0a + 1 Ob + -:2Vlo b ^cos(<P) (21)

where,

I_n A-0 ^■> Iθa A. Oa ¹ T Ob = ^Λ A 0b ^' and

φ = φa - φ_b

where, z=z_a-Z_}, lQ_a, and iQb are the intensities of the light propagating in the suspended waveguides. These intensities are functions of the gaps z_a, and Zf,, and in general, they are not the same. If for simplicity we consider only the first mode, then Iø_a, and In^ can be calculated from the area overlapping of the first mode, after going through the Gaussian expansion, with the first mode in the suspended waveguide. Assuming the light intensity propagating in the fixed waveguides is 1/2 then for example, for z_a=2 μm, and .l μm the light intensities entering the suspended waveguides are 0.76x1/2 and 0.81x1/2 respectively [Haronian 1998b].

The sensitivity of the GMI is:

where C = — ^-^

The intensity as a function of the displacement is drawn for two configurations in Figure 43. The first configuration is for z_a=z^=2 μm and in the second z_a=2 μm and z_j-,=1.7 μm. This figure shows that for an asymmetric configuration an AC input signal will result in a similar AC output signal. On the other hand, for a symmetric configuration the frequency of the output signal has twice the frequency of the input signal.

GM displacement sensors based on optical modulation in the form of energy losses and mode conversions, resulting from relative displacement of aligned and suspended waveguides are described in [Haronian 1998b] and hereinabove. For comparison with the GM displacement sensor the dashed line in Figure 43 is the intensity as a function of the displacement calculated for the first mode in a GM displacement sensor. C _j 1 in Figure 3 represent the coupling efficiency between the first mode in the transmitting waveguide and the first mode in the receiving waveguide. As will be further described below, this comparison shows the high sensitivity of the interferometry sensor.

Figure 44 shows the sensitivity of the GMI displacement sensor as a function of the displacement. The dashed line is the sensitivity as a function of the displacement of the GM sensor based on modes mismatch in a fixed- free waveguide [Haronian 1998b]. Both, the configurations of the GMI and the GM, are designed such that the zero displacement sensitivity is maximized. In the case of the interferometry sensor the maximum sensitivity can be approximately gained when z/λ =0.24. Figure 44 is drawn for z_a=2 μm, and z_j-=1.7 μm. In the case of the GM sensor the maximum sensitivity is gained when the sending waveguide and the receiving waveguides are misaligned by half of the waveguide width. Once again, this Figure shows the high sensitivity of the GMI displacement sensor relative to the GM displacement sensor. The general configuration shown in Figure 42 also illustrates that a

Fabry-Perot interferometer may form between the suspended waveguide and the merging waveguide. The transfer function of such FPI, that affects I_a, and I_j-„ is a function of the gap z_a, z_j-, and therefore, will be affected as these gaps are modulated. Nevertheless, this effect is expected to be relatively small as the expected finesse of the Fabry-Perot interferometer is low. Device configurations:

The GMI displacement sensor is fabricated using SCREAM (single crystal reactive ion etching and metalization) process [Shaw et al., 1994]. A description of fabrication process of the SOW displacement sensor technology is given above. As shown in Figure 45 the SOW displacement sensor are fabricated on top of a plateau formed by KOH etch. The distance between the input/output ports of the waveguides and the edge of the plateau is about 20 μm. The plateau is about 100 μm above the lower plane, which allows optical fibers with diameter smaller than 200 μm to approach the input/output ports of the waveguides.

Figures 46a-b are general views of a GMI based on the concept depicted in Figure 42. The central suspended frame supports two merging waveguides. Light is fed to the device through a Y fixed waveguide with cross section of 10 μm xlO μm. As shown in Figure 47 the fixed waveguides taper parabolically down to a cross section of 1.6 μm x 10 μm at the input to the suspended waveguides that are located about 2 μm away. It is found that such a parabolic transition between waveguides can be designed to operate adiabatically, where the lowest mode propagates through the parabolic section without transferring power to higher modes [Shaw et al., 1994; Haronian 1998b]. The merged section of the suspended waveguide feeds a fixed output waveguide with cross-section of 5 μm xlO μm. (see Figure 48). The output waveguide expands parabolically to cross section of 10 μm xlO μm and feeds at its other end a photodiode. Therefore, in addition to the optical signal resulting from the interferometry effect, the displacement 6 will lead to a signal due to the mismatch between the merging waveguide and the fixed waveguide that feeds the photodiode. Nevertheless as shown in Figures 43 and 44 this last signal is very small compared with the signal induced by the interferometry effect.

As shown in Figures 46a-b the supporting beams have an L shape and are connected to the frame close to its center rather than on its two sides. This was done in order to minimize the deformation of the device, after its release, as a result of intrinsic stress, that developed in the Siθ2, Si3N4 layer. From Figure 43 one sees that a small deformation leading to a vertical displacement of 0J μm of the frame is sufficient to drastically affect the input/output characteristics of the GMI displacement sensor. Experiments and results:

The sensor is excited acoustically and is tested with light with wavelength of 0.633 μm that is guided by the Si3 _t layer, and with light with wavelength of 1.3 μm that is guided by the silicon beam. Figure 49 is a schematic view of the testing setup. The input and output fibers have a 9 μm and a 50 μm core diameter respectively. The input fiber is coupled into the waveguide through a spherical lens and the output fiber is located next to the output port of the waveguide. The fibers and the sample are mutually manipulated using 5 degrees of freedom positioners. The output fiber is connected to a low noise photodiode, which in turn is connected to an oscilloscope or to a spectrum analyzer. The input fiber is chosen to have a 9 μm core diameter for both light sources. Therefore, in order to switch from one light source to the other only the light source and the photodiode are replaced. This leaves the input and output fibers aligned to the sensor. AC excitation:

Figure 50 is the response of the GMI displacement sensor to acoustical excitation that scans the 2-6 kHz range at 10 Hz. The test started by aligning the sensor to the optical fibers using the 0.633 μm light source under constant acoustical excitation. After the sensor was aligned and system resonance response was recorded, the 0.633 μm light source and the photodiode were replaced with the 1.3 μm light source and photodiode. At this stage no signal appeared at the photodiode. The signal appeared only after raising the GMI to allow the 1.3 μm light to be guided by the silicon waveguide.

As shown in Figure 43 for small displacements in an asymmetrical configuration the output signal has the same frequency as the frequency of the sensor vibration. The results of a Finite Element Analysis of the device, are described in Figure 51. This model show the first 4 vibrational modes of the device identifying the results in Figure 50 with the first vibrational mode of the device.

In both of the experiments described in Figure 50 a 1.5 mW diode laser was used. Therefore the results in Figure 50 indicates that signal recorded while using the silicon waveguide is much larger. In addition, it was much easier to align the 1.3 μm light source to the silicon waveguide. The reason for these findings is the cross sections dimensions of the input/output waveguides: The input/output cross section of the silicon waveguide is 10 μm xlO μm while that of the Si3N4 waveguide is 10 μm xθ.4 μm. Therefore the input/output coupling of the silicon waveguide are more efficient leading to a larger optical modulation. Impulse Excitation:

The sensor was excited acoustically at low frequency by applying a square waveform to the speaker at frequencies ranging between 1 Hz and

100 Hz. As the resonance frequency of the sensor is much higher than frequencies in this range a pulse response was observed. The sensor response to pulse excitation can be described as:

x(t) = Xc, exp[-ωJ/Q, ]cos( l - l/Q²ω J + φ₁ ) (23) where, the index refers to the vibrational mode, C_j is a constant, ω_j is the natural frequency of the mode i, Q_j is the quality factor, and φ_j is the phase. Figure 52 is the response of the sensor to acoustical impulse. In addition, a best fit, calculated from Eq. 5 is overlaid, with φ=0, ω=8000π sec"l. and Q=40π.

Thus, this Example describes a Geometrical Modulation based Interferometer (GMI) displacement sensor. The GMI sensor was fabricated using the suspended optical waveguide displacement sensors technology, which itself is based on the SCREAM process. The SCREAM process allows design of MEMS with an in-plane degree of freedom, and thus the SOW displacement sensor technology enables planer design of complex waveguide configurations. GMI performance were analyzed and its high sensitivity compared to the GM displacement sensor demonstrated. The sensor was tested using light sources with wavelength of 0.633 μm and 1.3 μm. The coupling to the silicon waveguide showed higher efficiency and the signal was about one order of magnitude stronger than that of the Si3N waveguide.

The performance of the GMI is sensitive to the exact lengths of the gaps z_a, and Z_{-. The SCREAM process, typically results in small residual stress that typically deflects the suspended mesh and thus change z_a, and z_jj, while keeping z_a+z_j- constant. One way to avoid this problem is to design a tunable GMI allowing a control over the actual working position of the mesh that supports the suspended waveguides. A byproduct of such tunability is the ability to control the sensitivity of the sensor. A tunable GMI displacement sensor is shown schematically in Figure 53, where the mesh is driven to the working position by the comb drive. Figure 54 is the calculated sensitivity as a function of z_a for z_a+z_j-=6 μm. The sensitivity is defined in Equation 22 and is shown in Figure 43. EXAMPLE 5 Direct integration (Dl) of solid state stress sensors with single crystal Micro-electro-mechanical systems for integrated displacement sensing

MEMS with planar degree of freedom show grate promise as they allow devices with variety of abilities. Still sensing displacements in the plane of the wafer is a complicated task as the sidewalls of the structures are not visible to the photolithography process. This paper introduces one solution to this disadvantage by Direct Integration (Dl) of stress sensors made of pn diodes and MOS transistors with MEMS made of micro-beams with planar degree of freedom. Micro-beams with typical cross section of 2 μm x 20 μm are fabricated from single crystal silicon using SCREAM (Single Crystal Reactive Ion Etching and Metalization) process [Shaw et al., 1994]. Sensors are integrated the support of the beams close to their roots. A finite element analysis shows that deformation of the beams induces stress that extends into the support of the beams. By modulating the band gap energy this stress affects the I-V transfer function of the sensor located at the support. This band gap modulation results in modulating secondary properties of solid state devices such as the intrinsic concentration, the charge carrier concentration, the built-in pn junction potential, the junction width, and the mobility of holes and electrons through the piezoresistance tensor.

In this paper the Dl concept is demonstrated by implementing planar accelerometers and flow sensors that are based on integration with pn diodes and with NMOS transistors. Accelerometers with sensitivity as high as 326 mv/g and flow sensor with amplitude and frequency sensitivities as high as 58 mv/ml/s and 250 Hz/ml/s are demonstrate. In-plane DOF vs. Out of plane DOF:

Planarity is a fundamental nature of microfabrication technology. This property is responsible to the fact that different element in VLSI such as diodes, transistors, capacitors etc., are placed and interact in the plane of the wafer. This planarity property is inherited also by the MEMS technology, and many MEMS are comprised of element such as sensors and actuators that are fabricated and interact in the plane of the wafer and therefore have an in-plane Degree Of Freedom (DOF) (see Fig. la). For example, the micro x-y-z stage carrying an STM tip [Xu et al., 1995], the vibrating gyroscope [Maenaka et al., 1996] or the micro-gear [Legtenberg et al., 1997] are all MEMS with several mechanical components interacting with each other in the plane of the wafer. In-spite of this planarity nature, MEMS with out-of-plane DOF can also be found. The micromachined microphone made of a suspended membrane over a sealed cavity [Yazdi and Najafi, 1997], and the pendulum accelerometer fabricated using wet etch of silicon [Bergqvist and Rudolf, 1994], are examples of devices with out-of-plane DOF. Still because of the planarity nature of the microfabrication technology these MEMS have low mechanical integration abilities and therefore are capable of performing only simple tasks.

The disadvantage of the planarity nature of the microfabrication technology becomes an advantage when it comes to integrating a displacement sensor for elements with out-of plane DOF. Since these elements move out-of the plane it is relatively easy to fabricate sensors such as capacitive sensors, piezoresistive, or piezoelectric sensors on these planar elements. The movement of the micromachined microphone for example, can be sensed capacitively by coating the membrane and a counter close plane with metal, or it can be sensed by coating piezoresistive or piezoelectric materials on the membrane. On the other hand, the advantage of the planarity nature of the microfabrication technology for elements with in-plane DOF becomes a disadvantage when it comes to integration of displacement sensors. Since elements with planar DOF move in the plane of the wafer the traditional sensing concepts should be fabricate on their sidewalls. Not only this task is not trivial with conventional microfabrication technology, typically the overall area of the sidewalls is too small to be effective. This is why comb structures, that increase the overall sidewalls area, are used for capacitive sensing

One can conclude that MEMS with out-of plane DOF have low mechanical integration capabilities and high sensing integration capabilities, while MEMS with in-plane DOF have high mechanical integration capabilities and low sensing integration capabilities. The disadvantages of the two technologies increase their cost and complexity. Therefore, in order for these technologies to become attractive their disadvantages should be eliminated. In order to increase the mechanical integration ability of a MEMS technology with out-of-plane DOF one need to fabricate suspended elements one on top of the other either using multiple chip technology or using non-planar microfabrication technology. Multiple chip technology is relatively expensive while non- planar microfabrication technology is not available yet. On the other hand in order to increase the sensing integration ability of a MEMS technology with in-plane DOF one need to associate the in- plane mechanical movement with some sensible physical property without adding fabrication complexity. The direct integration (Dl) technology does exactly that: It was found using a finite element analysis that the stress developing in a fixed-free single crystal beam during deformation extends into the support of the beam. The Dl uses this stress to modulate the electrical properties of a solid state sensors such as pn diodes, MOS or bipolar transistors integrated at the root of the beam. This Example describes the potential of this technology by demonstrating the performances of accelerometers and flow sensors with in-plane DOF that are integrated with pn diodes and NMOS transistors.

Stress affects the electrical properties of semiconductors by modulating the band gap energy. This affect was observed in the early 1960's by studying the performance of pn diodes and bipolar transistors [Wortman et al., 1964; 1966] under mechanical stress. The stress coefficient of the band gap energy was found to be in the order of 10" 12 eVcnvVdynes [Goetzberger and Finch, 1964]. This band gap energy modulation affects secondary properties of solid state devices. These secondary properties are the intrinsic concentration, the charge carrier concentration, the built-in pn junction potential, the junction width, and the mobility of holes and electrons through the piezoresistance tensor.

The stress distribution along the beam, when subject to bending moment is anti-symmetric. Therefore when the beam is displaced to the left, its left section is subject to compressive stress while the right section is subject to tensile stress. Assuming symmetrical electrical-stress relation, the electrical effects taking place in these sections will tend to cancel each other. This can be further understood if one considers the transfer function, f, of the sensor with relation to the configurations drawn in Figure 55. The sensing element integrated into the root of beam 'a', is placed such that it can be divide into two sub sensors with transfer functions fγ_Q=fl2, fχo fll connected in serial. When the beam bends the transfer function of these sub-sensors change such that /, = — ±— , /, = — +— . Therefore the ^to ' 2 2 2 2 transfer function after stress is applied is . = , + /- = ± JΛ — i _? and the effective change is /. - / = ± ' ^~ . This change is much smaller than the change that would be observed if the sensor was uniformly stressed. In addition, if the properties of the sensor is homogeneous and if the stress is uniformly distributed then Af = Δ/₂ and the electrical effects generated by the stress will not be observed.

The sensor integrated into beam 'b' in Figure 55 is oriented such that it can be divide into two sub sensors with transfer functions fγ _Q ⁼2f, f _Q^f connected in parallel. If one assumes that the sensor is homogeneous and if the stress is uniformly distributed then when the beam bends the transfer functions change such that /, = if ±_bf, f₂ = 2f+2Af . Therefore, the transfer function after the stress was applied is/, = f_l '₊f ² ₂ = /-— f— and the effective

change is - One can see that although the electrical effect

generated by the stress is observed it is much smaller than if the sensor was uniformly stressed.

The first configuration (beam a) is valid for the MOS transistor sensor, and the second configuration (beam b) is valid for the pn diode sensor. One concludes that for symmetrical configurations the electrical effects that are excited by stress are smaller than those that can be expected if a uniform stress is applied on the sensors. In order to reduce this effect it is necessary to place the sensor off-axis as much as possible. Such off-axis placement will increase the initial difference between the two sub sensors and will increase the electrical response.

In addition, it is possible to take for advantage the anti-symmetric stress effect, by fabricating two stress sensors one next to the other symmetrically on the support. In this case the transfer function of the two sensors is R and if instead of summing the electrical signal they are fed into a differential amplifier the electrical effect will be doubled. Integration of fabrication processes:

Wafers with IC were initially designed and fabricated in a commercial VLSI facility. This fabrication included the pn diodes and NMOS transistors integrated with IC that amplifies and filters the signals. Next the wafers went through ICP based SCREAM process and the root of the beams of the designed devices were aligned to their sensors. The ICP SCREAM process uses the Bosch process instead of the chlorine etch, and thus allows the fabrication of structures with larger aspect ratio. Figure 56a is a scanning electron micrograph showing a single beam integrate with two diodes fabricated on the support of the beam symmetrically. The two p⁺⁺ notations represent the conductors that are connected to the two p⁺⁺ regions of the diodes while the single 'n' notation represent the conductor that is connected to the common 'n' region of the diodes. The common 'n' region is grounded while the p⁺⁺ regions are connected to a current source and to a differential amplifier. When the beam deforms to one side a compressive stress develops in this side while a tensile stress develops in the other side. Therefore, opposite stress effects are expected at the two diodes. These effects change the voltage drop applied on the diodes, and their difference is amplified by on chip differential amplifier. Figure 56b is a scanning electron micrograph showing a single beam integrated with an NMOS transistor. The S, D, G notations represent the source drain and gate of the transistor. The source of this transistor together with the source of an identical transistor, located outside the stressed area, are connected in parallel to a current source, their drains are fed into a differential amplifier, and their gates are held at the same potential. In this configuration the difference in the current flowing through the channel of the two transistors is amplified by the differential amplifier. Note that the stressed transistor is located slightly off-axis. This is done in order to reduce opposite electrical effects as a result of tensile and compressive stresses that are developing in the two sides of the beam. Figures 56c-d are SEM pictures showing two beams with common support integrated with two NMOS transistors. In Figure 56c two parallel beams are connected to the transistors such that each beam is connected to the channel of each transistor, and in Figure 56d, two vertical beams are connected to the transistors such that the channels of the two transistors are located at the edges of the beam. The sources of the two transistors are connected in parallel to a current source, their drains are fed into a differential amplifier, and their gates are held at the same potential. Therefore this measuring configuration measures the difference in the current flowing in the channels of the two transistors.

Figures 57a-b show a proof mass supported by four beams. One of the beams is integrated with NMOS transistor. The proof mass is about 1.7 x 10^~8 kg and the overall spring constant of the beams is about 4.6 N/m. The calculated resonance frequency of this lumped system is found to be 2.6 kHz. Figure 57c shows the response of the sensor to acceleration of 0.03 g at different frequencies. Figure 4d shows the response of the sensor to acceleration at different amplitudes in atmospheric pressure. The calculated off resonance sensitivity is found to be 0.8 mv/g.

Figure 58a shows an L shaped sensor with proof mass at its end, integrated with two diodes. The proof mass is 2.7 x 10^~10 kg and the first bending mode of the beam has a spring constant of 0.2 N/m. The calculated resonance is found to be A A kHz. Figure 58b shows the response of the sensor to acceleration of 0.03 g at different frequencies and different pressures revealing it resonance characteristic. The measured Q factor at 1.8xl0"3 bar is calculated to be about 400. Figure 58c shows the response of the sensor to acceleration at different amplitudes in atmospheric pressure. The calculated off resonance sensitivity is found to be 326 mv/g. Figure 58d shows the response of the sensor to acoustical and mechanical shocks. A similar resonance characteristics is excited by these shocks, that is slightly shifted to lower frequencies.

A little more complicated configuration is shown in Figures 59a-b. Here a cantilevered coiled beam with proof mass at its end is integrated with a single NMOS transistor and with two pn diodes. Figure 59c shows the response of the sensor integrated with the NMOS transistor to acceleration of 0.03 g from four directions. These result shows that while the sensors described in Figures 57a-b and 58a are sensitive to one direction this sensor is sensitive to one plane.

The L-shaped sensor and the coiled sensor were also excited with nitrogen flow. This flow forces the sensor to resonate at its natural frequency, and therefore similar to the floating element flow sensor [Roche et al., 1996; Su and Evans, 1996], this sensing concept is based on the interaction of the flow with the mechanical properties of the sensors. Nevertheless, unlike the floating element sensor, this sensing concept is based on resonance response rather than on deformation. Potentially, resonance response may have higher resolution as the noise floor is typically lower. Figure 59d is the response of the coiled sensor to flow at different flow rates. The calculated amplitude and frequency sensitivities are found to be 58 mv/ml/s and 250 Hz/ml/s respectively. Although the devices described above are simple they demonstrate the abilities of the Dl technology. This technology can be easily integrated into microactuators. One such example is shown in Figures 60a-b where an NMOS sensor is integrated with one of the bending beams that support a comb-drive actuator.

The direct integration technology was demonstrated by integration with in-plane DOF devices that were fabricated with SCREAM process. Accelerometers with sensitivity as high as 326 mv/g and flow sensor with amplitude and frequency sensitivity as high as 58 mv/ml/s and 250 Hz/ml/s were demonstrated. The mechanical part of the sensors were fabricated using SCREAM process and therefore this fabrication step adds only one lithography step to the overall fabrication process. This integration is therefore very efficient and cheap. Although the Dl concept is demonstrated using the SCREAM process, it may also be applied to other microfabrication technology such as those that are based on SOI.

EXAMPLE 6 Vibrating gyroscope based on the suspended optical waveguide (SWO) displacement sensors technology or on using the direct integration (Dl) technology

A vibrating gyroscope based on the SOW displacement sensor and Dl technologies is described in this Example. These devices use Coriolis force to sense rotation rate as low as 0.001 degree/sec. Such micromechanical devices are currently being developed by others using capacitance sensing. Generally these latter sensors are not sensitive enough for fine navigation and large effort is being made to increase their sensitivity. The gyroscope described herein is based on optical sensing of the Coriolis force and is therefore highly sensitive.

Angular rate sensors are used in automobile and avionics industries. Conventional sensors are based on gyroscopes that can sense the change in the rotation of a body with accuracy of up to 0.001 degree/sec. Still, these conventional gyroscopes are extremely expansive and thus their use is not fully exploit. In the past years, several vibrating micromachined gyroscopes and rate sensors were demonstrated. These sensors are based on vibration of mass that creates a Coriolis force with a magnitude that depends on their rotation rate.

The Ballerina effect or Coriolis force:

Figure 61 demonstrates the effect of Coriolis force on the rotation of a Ballerina. As is well known, when a ballerina rotates and moves here hand in and out, here speed of rotation changes. When the hands are moved closer to the body the rotation rate increases in order to maintain a constant moment of inertia. This means that a force, called Coriolis force, is applied such that rotation rate increases. In a similar way, when the ballerina moves here hands away from here body, the rotation rate of her body decreases for the same reason.

The conventional micromachined gyroscope: Figure 62 shows how this effect is used to sense the rotation rate in a conventional vibrating gyroscope. The large (proof) mass is electrostatically forced to vibrate at frequency fγ using a comb drive. This is equivalent to the movement of the ballerina hand in and out very rapidly at a frequency of f γ . In addition the gyroscope is rotating at frequency w around a given axis. This is equivalent to the rotation of the ballerina. Therefore a vibrating force is applied on the large mass that, depending on the direction of the large mass, is directed either in the direction of the rotation (When the mass moves closer to the axis of rotation) or in the opposite direction of the rotation (when the mass moves away form the axis of rotation). This force bends the springs that support the large mass such that the mass either approaches the wafer plane or draws away from it. This results in a change in the capacitance between the large mass and the plane of the wafer, and this capacitance change is a measure to the Coriolis force and thus to the rotation rate.

Advantages and disadvantages in MEMS technology: Before discussing the proposed gyroscope, it is essential to understand the current optical technology from the point of view of MEMS. Optical sensing is probably the most sensitive sensing technology. For example, optical manipulation in near field is used to reconstruct surfaces down to atomic level. Fabry-Perot or Mach-Zehnder, displacement sensors with Angstrom resolution can be configured. Nevertheless, optical sensors are rare in MEMS. Consider as an example the micromachined vibrating gyroscopes. These gyroscopes are based on Coriolis forces and often required to sense forces as small as 10^~ 5 N. Still, most of the effort by the major players in this field is directed to capacitive sensing that is by nature far less sensitive than optical sensing. Optics is not common in MEMS because of the leak in the ability to, cost effectively, integrate optical components such as light sources and sensors in silicon micromachining. The capacitive vibrating gyroscopes mentioned above can be microfabricated using standard microelectronics and MEMS technologies. These technologies maintain the most important characteristics of microelectronics technology namely mass production and high yield. These characteristics are not available with common opto-MEMS technology mostly because there is a leak in appropriate microfabrication technology that allows integrative optical sensor in silicon. Almost all of the optical elements that are used to translate mechanical effects into optical phenomena have out of plane degree of freedom, while most of the MEMS technology is based on in plane degree of freedom. About two yeas ago the SOW displacement sensor (suspended optical waveguide displacement sensors) technology was developed [Haronian, 1998a-d]. This technology uses suspended optical waveguide with in-plane degree of freedom to translate mechanical effects into optical modulations. Several sensors were fabricated and tested, among them are tips interacting in near field, and interferometery sensor, which are further described hereinabove. The SOW displacement sensor technology, because of its in-plane degree of freedom, opens new ways to integrate optical sensors with MEMS not only as micro-sensors per se, but also as a sensing mean for micro-actuators.

A light source which can be used with a Si3N4 waveguide is in the range 0.8 μm. It is possible to fabricate a PIN diode on the chip right at the end of the exit waveguide. It is also possible to use a silicon LED, electroluminescence polymers, or Si-GaAs technology. The gyroscope:

In the micromachined gyroscope according to the present invention the rotation is peφendicular to the plane of the wafer, and the mass vibrates in the plane of the wafer by bending horizontal beams 200 as shown in Figures 63 and 64. When the rotation rate Ω changes the Coriolis force changes leading to the bending of vertical beams 200. This bending is sensed optically using the SOW displacement sensor technology or by stress sensor at the root of the vertical beams which is further described below, using the Dl technology.

Direct integration with optics: As further shown in Figures 63-64 the left vertical beam 200 is also a waveguide. This waveguide-beam guides light from a light source 202 to a tapered edge. Light emerges from the tapered end and enters a fixed waveguide 204 that guides the light to a photodiode 206. When this waveguide-beam bends, the coupling between the two waveguides changes and the light intensity is modulated. This modulation server as a measure to the change in the Coriolis force and to the change in the rotation rate. It was calculated that such tip coupling taking place in near filed can be extremely sensitive. The resolution is estimated to be in the range of 1 Angstrom. In addition to the tip-tip sensor, additional SOW displacement sensor based sensor can be integrated in the gyroscope. One such sensor is the interferometer sensor [Haronian 1998c]. Direct integration with VLSI:

The Dl technology is described under Example 5 above, in Haronian 1999, and in U.S. Pat. application No. 09/101,014 and in PCT/IL96/00190, all of which are incoφorated herein by reference. In Figure 65, the two vertical beams are integrated with a pn diodes or transistors 210. The resolution of the L shaped cantilevered beam with proof mass at its end and a double pn diode sensor at its root is estimated to be about 100 angstrom. The sensitivity of the Dl technology is lower than that of most of the sensors based on the SOW displacement sensor technology.

Although the invention has been described in conjunction with specific embodiments thereof, it is evident that many alternatives, modifications and variations will be apparent to those skilled in the art. Accordingly, it is intended to embrace all such alternatives, modifications and variations that fall within the spirit and broad scope of the appended claims. All publications cited herein are incoφorated by reference in their entirety.

REFERENCES

1. Asif-Khan M., et al. Mat. Res. Soc. Proc. Vol. 339 (1994).

2. Benaissa K., A. Nathan, IC Compatible Optomechanical Pressure Sensors Using Mach-Zehnder Interferometery, IEEE Transactions on Electron Devices, 43 (9), 1571-1582, (1996)

3. Bergqvist J., Rudolf F. (1994) A silicon condenser microphone using bond and etch-back technology. Sensors and Actuators A 45:115-124.

4. Betzig E. and J. K. Trautman, Science 257, 189 (1992).

5. Bosselman T. and R. Ulrich, in " Proc. 2nd Int. Conf. On Fiber Sensors, Stutgart 1984, pp. 321-338.

6. Burcham K. E., G. N. De Brabander, J. T. Boyd, Micromachined Silicon Cantilever Beam Accelerometer Incoφorating an Integrated Optical Waveguide, SPIE, 1793, Integrated optics and Microstructures, 12-18, (1992)

7. Burroughs J. H., D. D. C. Bradley, A. R. Brown, R. N. Marks, K. Mackay, R. H. Friend, P. L. Burns and A. B. Holmes, Nature. 1990, 347, 539.

8. Chayet H., R. Pogreb and D Davidov, Phys. Rev. B. 56, 12702 (1997); Proceedings of the SPI 3148, California, USA (1997), p.34)

9. Cline J A, Isaacson M, 1995 Probe-sample interactions in reflection near-field scanning optical microscopy, Applied Optics, 34, (22), 4869- 76.

10. Davis R. F., Proc. IEEE 79, 702 (1991) l l. Degani O., D. J. Seter, E. Socher, S. Kaldor, Y. Nemirovsky, "Optimal design and noise consideration of micromachined vibrating rate gyroscope with modulated integrative differential optical sensing," Journal of Microelectromechanical Systems, 7, 329-338, (1998)

12. Diaz-Garsia M. A., F. Hide, B. J. Schwartz, M. D. McGehee, M. R. Anderson and A. J. Heeger, Appl. Phys. Lett. 1997, 70, 3191.

13. Eng T. T. H., S. C. Kan, G. K. L. Wong, Surface-micromachined epitaxial silicon cantilevers as movable optical waveguides on silicon- on-insulator substrates, Sensors and Actuators A, 49, 109-113, (1995)

14. Fang S. F., K. Adomi, S. Iyer, H. Morcoc, H. Zabel, C. Choi, and N. Otsuka, J. Appl. Phys. 68, R31, (1990).

15. Frolov S. V., A. Fujii, D. Chinn, Z. V. Vardeny, K. Yoshino and R. V. Gregory, Phys. Rev. Lett. 1998, 72, 2811. lό. Garthe D 1993 Fiber- and integrated-optical microphones based on intensity modulation by beam deflection at a moving membrane, Sensors and Actuators A, 37-38, 484-8.

17. Goetzberger A., R.H. Finch, "Lowering the breakdown voltage of p-n junction by stress", J. Appl. Phys., 35(6), 1851-1854, (1964)

18. Guha S., et al, Appl. Phys. Lett. 72, 415 (1998a).

19. Guha S., et al. Appl. Phys. Lett. 73, 1487 (1998b).

20. Haronian D., "Suspended optical waveguide with in-plane degree of freedom for microelectromechanical applications," Electronics Letters, 7, 34, pp. 663-664 (1998a)

21. Haronian D., "In-plane degree of freedom optical waveguide displacement sensors based on geometrical modulation," Sensors and Actuators A, 69. pp. 217-225 (1998b).

22. Haronian D., "Geometrical Modulation based interferometry for displacement sensing using optically coupled suspended waveguides," Journal of Microelectromechanical systems, 7 (3), pp. 309-314, (1998c).

23. Haronian D. "Displacement sensing using geometrical modulation in reflection mode (GM-RM) of coupled optical waveguides ' Journal of micromachining and microengineering, 8:323-326 (1998d).

24. Haronian D., "Direct Integration (Dl) of solid state stress Sensors with Single Crystal Micro-electro-mechanical systems for integrated displacement sensing," Twelfth IEEE International Micro Electro Mechanical Systems Conference January 17-21, 1999, pp. 88-93. 1999 Buena Vista Palace Resort Orlando, Florida USA.

25. Hunspurger R. G., "Integrated Optics Theory and Technology" Springer, Berlin, 1982

26. Jones B. E. and J. S. McKenzie, Sensors and Actuators A 37-38, 202 (1993). M. Tabib-Azar "Integrated Optics and Microstructure Sensors" Kluwer, Boston, 1995.

27. Kim Y. W., M. G. Allen, N. F. Hartman, Micromechanically based integrated optic modulators and switches, SPIE Vol, 1793, 183-198 (1992).

28. Legtenberg R., Berenschot E., Elwenspoek M., Fluitman J. H. (1997) A fabrication process for electrostatic microactuator with integrated gear linkages. Journal of Microelectromechanical Systems 6:234-241.

29. Maenaka K., Fujita T., Konishi Y., Maeda M. (1996) Analysis of a highly sensitive silicon gyroscope with cantilever beam as a vibrating mass. Sensors and Actuators A 54: 568-573. 30. Marcatili E. A. J., Dielectric Rectangular Waveguide and Directional

Coupler for Integrated Optics, The Bell System Technical Journal, 2071-

2101, Sept. (1969) 31. Milton A. F., W. K. Burns, Mode coupling in optical waveguide horns,

QE-13, 10, 828-835, (1977). 32. Nabor F., B. Kippelen, E. A. Mash, P. M. Allemand, M. Kuwata-

Gonokami, K. Takeda and N. Peyghambarian, Appl. Phys. Lett. 1998,

72, 141. 33. Nakamura S., et al. Appl. Phys. Lett. 62, 2390 (1993) 34. Nakamura S., et al. Jpn. J. Appl. Phys. 35, L74 (1996). 35. Nakamura S. and G. Fasol "The Blue Laser Diode" Springer, Berlin,

1997. 36.Noach S., E. Earaggi , C. B. Cohen Y Avny, R. Neumann, D. Davidov and A. Lewis , Appl. Phys. Lett. 69, 3650 (1996). 37. Paternottre A, Molpeceres F, Montero de Espinosa F R, Carbo R,

Chinchurreta F J, Muriel M A 1993 Acoustic-field fiber-optic sensor,

Sensors and Actuators A, 37-38, 489-93. 38.Pohl D. W., in "Advances in Optical and Electron Microscopy" C. R.

Sheppard and T. Mulvey, Eds (Academic, London, 1990).

39. Roche D., C. Richard, L. Eyrud, C. Audoly, "Piezoelectric bimoφh bending sensor for shear-stress measurement in fluid flow, "Sensors and Actuators A, 55, 157-162, (1996)

40. Salzman J. and H. Temkin, Mater. Sci. & Engineer. B50, 148 (1997). 41. Sawate'ev V., A.V. Yakimov, D. Davidov , R. M. Pogreb, R. Neumann and Y Avny, Appl. Phys. Lett. 71,3344 (1997).

42. Shaw K. A., Z. L. Zhang, N. C. MacDonald, SCREAM 1: A Single Mask, Single-Crystal Silicon, reactive ion etching process for microelectromechanical structure, Sensors and Actuators A, 40, 63-70, (1994).

43. Soref A., J. P. Lorenzo, All-silicon active and passive guided-wave components for λ=1.3 and 1.6 μm, IEEE J. Quantum Electronics, QE- 22, 873-879, (1986)

44. Strite S. and H. Morkoc, J. Vac. Sci. Technol. B10, 1237 (1992).

45. Stutius W., W. Streifer, Silicon nitride films on silicon for optical waveguides, Applied Optics, 16, 12, 3218-3222, 1977

46. Su Y., A. G. R. Evans, A. Brunnschweiler, "Micromachined silicon cantilever paddles with piezoresistive readout for flow sensing," J. Micromech. Microeng. 6. 69-72, (1996) 47. Sze S. M.; "Semiconductor Sensors", (Wiley, New York, 1994).

48. Tabib-Azar M., "Microactuators, Electrical, Magnetic, Thermal, Optical, Mechanical, Chemical & Smart Structures" Kluwer, Boston, 1998.

49. Tamir T., "Integrated Optics" Springer-Verlag, New York, 1982.

50. Tamir T., in "Integrated Optics" T. Tamir (Editor), Springer, Berlin 1979.

51. Tessler N., G. J. Denton and R. H. Friend, Nature, 382, 695 (1996) 52. Tessler N., et al. Adv. Mater. (1998)

53. Thomson W T, 1971 Vibration theory and applications, pg. 39 (Prentice-Hill Inc., USA)

54. Tisch U., et al. "Temperature Dependence of Optical Constants in GaN" Europ. Mater. Res. Soc, Strasbourg, June 16-29 (1998a).

55. Tisch U., et al. " Surface Moφhology of GaNO." ICCG 12-ICVGE10, Jerusalem, July 26-31 (1998b).

56. Tran A.T.T.D., D. Haronian, N. A. Switz, Y-H. Lo, N. MacDonald, "Micromachined Tunable Micro Interferometer for Optical Accelerometer", CLEO '95, Baltimore, Maryland, USA, Postdeadline, CPD18, May 22-26, 1995.

57. Tran A.T.T.D., Y.H. Lo, Z.H. Zhu, D. Haronian, E. Mozdy "Surface micromachined Fabry-Perot tunable filter". IEEE Journal of Photonics Technology Letters, 8 (3), 393-395, 1996.

58. Wortman J. J., J.R. Hauser, R.M. Burger, "Effect of Mechanical Stress on p-n Junction Device Characteristics", J. Appl. Phys., 35(7), 2122- 2131, (1964)

59. Wortman J. J., J.R. Hauser, "Effect of Mechanical Stress on p-n Junction Device Characteristics, II. Generation-Recombinant Current", J. Appl. Phys., 37(9), 3527-3530, (1966)

60. Wu S., H. J. Frankena, Integrated optical sensors using micromechanical bridges and cantilevers, SPIE, 1793, Integrated optics and Microstructures, 83-89, (1992)

61. Xu Y., MacDonald N. C., Miller S. A. (1995) Integrated micro-scanning tunneling microscope. Applied Physics Letters 67: 2305-2307. 62. Yazdi N., Najafi K. (1997) An all-silicon single- wafer fabrication technology for precision microaccelerometer. In: Tech. Dig. 9th Int.

Conf. Solid-state sensors and actuators (Transducers '97) Chicago IL pp

1181-1184.

Claims

WHAT IS CLAIMED IS:

1. A micromachined displacement sensor chip comprising:

(a) a reference frame;

(b) at least one suspended waveguide element having an in-plane degree of freedom being integrally formed with said reference frame;

(c) a light source being integrally formed with said reference frame and being optically coupled to said at least one suspended waveguide element at one end thereof; and

(d) a light sensor being integrally formed with said reference frame and being optically coupled to said at least one suspended waveguide element at another end thereof; such that when said at least one suspended waveguide element is subjected to an external force, an in-plane displacement of said at least one suspended waveguide element is monitorable by said light sensor due to light modulation.

2. The micromachined displacement sensor chip of claim 1, further comprising at least one suspended proof mass integrally formed with said at least one suspended waveguide element.

3. The micromachined displacement sensor chip of claim 1, wherein said at least one suspended waveguide element also serves as a suspended proof mass.

4. The micromachined displacement sensor chip of claim 1, further comprising at least one fixed waveguide element integrally formed with said reference frame, said at least one fixed waveguide is optically coupled through one end thereof to said light source and through another end thereof to said at least one suspended waveguide element, thereby optically coupling between said at least one suspended waveguide element and said light source.

5. The micromachined displacement sensor chip of claim 4, wherein said at least one fixed waveguide element and said at least one suspended waveguide element are optically coupled via a tip-tip optical coupling.

6. The micromachined displacement sensor chip of claim 4, wherein said at least one fixed waveguide element and said at least one suspended waveguide element are optically coupled via a tip-blunt end optical coupling.

7. The micromachined displacement sensor chip of claim 4, wherein said at least one fixed waveguide element and said at least one suspended waveguide element are optically coupled via a blunt end-tip optical coupling.

8. The micromachined displacement sensor chip of claim 4, wherein said at least one fixed waveguide element and said at least one suspended waveguide element are optically coupled via a reflector to effect geometrical modulation in reflection mode.

9. The micromachined displacement sensor chip of claim 4, wherein said at least one fixed waveguide element includes a splitting fixed waveguide element and further wherein said at least one suspended waveguide element includes a combining suspended waveguide element, said splitting fixed waveguide element and said combining suspended waveguide element are optically coupled such that light arriving from said light source and guided through said splitting fixed waveguide element recombines in said combining suspended waveguide element to thereby form an interferometer.

10. The micromachined displacement sensor chip of claim 1, further comprising at least one fixed waveguide element integrally formed with said reference frame, said at least one fixed waveguide is optically coupled through one end thereof to said light sensor and through another end thereof to said at least one suspended waveguide element, thereby optically coupling between said at least one suspended waveguide element and said light sensor.

1 1. The micromachined displacement sensor chip of claim 10, wherein said at least one fixed waveguide element and said at least one suspended waveguide element are optically coupled via a tip-tip optical coupling.

12. The micromachined displacement sensor chip of claim 10, wherein said at least one fixed waveguide element and said at least one suspended waveguide element are optically coupled via a tip-blunt end optical coupling.

13. The micromachined displacement sensor chip of claim 10, wherein said at least one fixed waveguide element and said at least one suspended waveguide element are optically coupled via a blunt end-tip optical coupling.

14. The micromachined displacement sensor chip of claim 10, wherein said at least one fixed waveguide element and said at least one suspended waveguide element are optically coupled via a reflector to effect geometrical modulation in reflection mode.

15. The micromachined displacement sensor chip of claim 1, wherein said force is an acceleration force, the micromachined displacement sensor chip serves as a micromachined accelerometer chip.

16. The micromachined displacement sensor chip of claim 1, wherein said force is a Coriolis force, the micromachined displacement sensor chip serves as a micromachined gyroscope chip.

17. The micromachined displacement sensor chip of claim 2, wherein said force is a Coriolis force, the micromachined displacement sensor chip further includes an electrostatic actuator integrally formed with said reference frame for actuating said proof mass in plane vertically to said Coriolis force, the micromachined displacement sensor chip serves as a micromachined gyroscope chip.

18. The micromachined displacement sensor chip of claim 1, wherein said force is selected from the group consisting of thermal expansion force, electrostatic force, magnetic force and piezoelectric force, whereas said suspended waveguide element is selected responsive to said thermal expansion force, said electrostatic force, said magnetic force and said piezoelectric force, respectively.

19. A micromachined displacement sensor chip comprising:

(a) a reference frame;

(b) a first waveguide element being integrally formed with said reference frame;

(c) a second waveguide element being integrally formed with said reference frame;

(e) a light source being integrally formed with said reference frame and being optically coupled to said first waveguide element at one end thereof; and

(f) a light sensor being integrally formed with said reference frame and being optically coupled to said second waveguide element at one end thereof;

(g) a reflector integrally formed with said reference frame and optically coupling said first waveguide element with said second waveguide element; wherein at least one of said reflector, said first waveguide element and said second waveguide element serves as a suspended proof mass to effect geometrical modulation in reflection mode when displaced as a response to an external force exerted thereon.

20. A micromachined chip comprising:

(a) a reference frame;

(b) at least one suspended waveguide element having an in-plane degree of freedom being integrally formed with said reference frame; and

(c) a light source being integrally formed with said reference frame and being optically coupled to said at least one suspended waveguide element at one end thereof.

21. A micromachined chip comprising:

(a) a reference frame;

(c) a light sensor being integrally formed with said reference frame and being optically coupled to said at least one suspended waveguide element at one end thereof.

22. A micromachined displacement sensor chip comprising:

(a) a reference frame;

(b) at least one suspended element having an in-plane degree of freedom being integrally formed with said reference frame through a root thereof;

(c) a solid state sensor being integrally formed in said root, such that when said at least one suspended element is subjected to an external force, an in-plane displacement of said at least one suspended element is monitorable by said solid state sensor.

23. The micromachined displacement sensor chip of claim 22, further comprising at least one suspended proof mass integrally formed with said at least one suspended element.

24. The micromachined displacement sensor chip of claim 22, wherein said at least one suspended element also serves as a suspended proof mass.

25. The micromachined displacement sensor chip of claim 22, wherein said force is an acceleration force, the micromachined displacement sensor chip serves as a micromachined accelerometer chip.

26. The micromachined displacement sensor chip of claim 22, wherein said force is a Coriolis force, the micromachined displacement sensor chip serves as a micromachined gyroscope chip.

27. The micromachined displacement sensor chip of claim 23, wherein said force is a Coriolis force, the micromachined displacement sensor chip further includes an electrostatic actuator integrally formed with said reference frame for actuating said proof mass in plane vertically to said Coriolis force, the micromachined displacement sensor chip serves as a micromachined gyroscope chip.

28. The micromachined displacement sensor chip of claim 22, wherein said force is selected from the group consisting of thermal expansion force, electrostatic force, magnetic force and piezoelectric force, whereas said suspended element is selected responsive to said thermal expansion force, said electrostatic force, said magnetic force and said piezoelectric force, respectively.