Integrated, microminiature electric to fluidic valve and pressure/flow regulator

What is claimed is:

1. An integrated pressure regulator comprising:

a first integrated valve comprising:

a first wafer;

a second wafer having a first surface and a second surface and having a trench etched into said first surface deep enough to leave a flexible diaphragm of the material of said second wafer defined by the bottom wall of said trench and said second surface, said silicon wafer being bonded to said first wafer such that said trench defines a sealed cavity;

a material trapped in said cavity which has a boiling point above the highest ambient temperature that might be experienced;

means for increasing the temperature of said trapped material to said boiling point so as to move said diaphragm;

and wherein said first wafer is pyrex and said second wafer if silicon and further comprising a second pyrex wafer having a first and second surface, said first surface having an input channel and an output channel etched therein, said input channel and said output channel separated by a valve seat in the form of a wall separating said input channel from said output channel, said first surface being bonded to said second surface of said silicon wafer at substantially all points except a predetermined area surrounding where said diaphragm mates with said valve seat;

and wherein said input channel and said output channel each have a plurality of projecting fingers which are interdigitated and wherein said valve seat takes the form of a serpentine mesa which separates the fingers of the input channel from the output channel fingers;

and wherein said means for heating is a resistive material deposited on the surface of said first pyrex wafer which seals said trench to form said cavity so as to be included within the perimeter of said cavity such that current may be passed through said resistive material to heat the fluid in the cavity

wherein said first and second pyrex wafers are anodically bonded to said silicon wafer and wherein the surface of said valve seat which contacts said diaphragm is coated with a material which prevents anodic bonding between said diaphragm and said valve seat;

said first integrated valve having its input port for coupling to a high pressure source;

a second integrated valve comprising:

a first wafer;

a second wafer having a first surface and a second surface and having a trench etched into said first surface deep enough to leave a flexible diaphragm of the material of said second wafer defined by the bottom wall of said trench and said second surface, said silicon wafer being bonded to said first wafer such that said trench defines a sealed cavity;

a material trapped in said cavity which has a boiling point above the highest ambient temperature that might be experienced;

means for increasing the temperature of said trapped material to said boiling point so as to move said diaphragm;

and wherein said first wafer is pyrex and said second wafer is silicon and further comprising a second pyrex wafer having a first and second surface, said first surface having an input channel and an output channel etched therein, said input channel and said output channel separated by a valve seat in the form of a wall separating said input channel from said output channel, said first surface being bonded to said second surface of said silicon wafer at substantially all points except a predetermined area surrounding where said diaphragm mates with said valve seat;

and wherein said input channel and said output channel each have a plurality of projecting fingers which are interdigitated and wherein said valve seat takes the form of a serpentine mesa which separates the fingers of the input channel from the output channel fingers;

and wherein said means for heating is a resistive material deposited on the surface of said first pyrex wafer which seals said trench to form said cavity so as to be included within the perimeter of said cavity such that current may be passed through said resistive material to heat the fluid in the cavity

wherein said first and second pyrex wafers are anodically bonded to said silicon wafer and wherein the surface of said valve seat which contacts said diaphragm is coated with a material which prevents anodic bonding between said diaphragm and said valve seat;

said second integrated valve sharing the same silicon wafer and the same pyrex wafers as said first integrated valve and having said input port of said second integrated valve for coupling to a low pressure effluent sink;

an output port for coupling to said output channels of both said first and said second integrated valves; and

a pressure sensor coupled to said output port.

2. An apparatus as defined in claim 1 wherein said pressure sensor comprises a diaphragm formed in the silicon wafer shared by said first and second integrated valves, said diaphragm in fluid communication with said output port and forming an evacuated chamber between said diaphragm and said first pyrex wafer shared by said first and second integrated valves, said evacuated chamber having a first capacitor plate formed by a conductive coating on at least a portion of the surface of said first pyrex wafer within said cavity and having a second capacitor plate formed by a conductive coating on at least a portion of the surface of said diaphragm.

3. An integrated pressure regulator comprising:

a first wafer having first and second surfaces, said second surface having first and second resistor patterns formed thereon and having a first conductive film formed thereon;

a second wafer bonded to said first wafer and having first and second surfaces, said first surface having first and second trenches formed therein, said trenches having a depth sufficient to form first and second flexible diaphragms between the bottoms of said trenches and said second surface, and having a third trench formed in said first surface, the bottom of said third trench being covered with a second conductive film, and having a fourth trench formed in the second surface opposite said fist surface, said fourth trench formed deep enough that a flexible diaphragm is formed between the bottoms of said third and fourth trenches, said first and second trenches being formed at locations aligned with the locations of said first and second resistor patterns such that said patterns are within the cavities formed between said first and second wafers at the locations of said first and second trenches, said third trench being located such that when the first and second wafers are bonded together in a vacuum, an evacuated cavity with said first and second conductive films forming a capacitor with a vacuum dielectric is formed; and

a third wafer having first and second surfaces, said first surface being bonded to said second surface of said second wafer at predetermined locations and having a first pair of input and output channels formed in said first surface and separated by a first valve seat in the form of a wall between said input and output channels which said first diaphragm contacts but which is not bonded to said first diaphragm, said first surface also having a second pair of input and output channels separated by a second valve seat in the form of a wall between said input and output channels which said second diaphragm contacts but which is not bonded to said second diaphragm, said second surface having a hole formed therein passing through said third wafer to emerge from said first surface at a location which is fluid communication with said fourth trench.

4. An integrated pressure regulator comprising: first and second integrated valves comprising:

a first semiconductor wafer having first and second parallel and opposite surfaces and having a first trench formed in said first surface so as to leave a first flexible diaphragm between the bottom of said trench and said second surface of said wafer and having a second trench formed in said first surface so as to leave a second flexible diaphragm between the bottom of said second trench and said second surface of said wafer, and having third and fourth trenches formed in said first and second surfaces, respectively, so as to form a diaphragm, the bottom of said third trench being coated with a conductive film;

a second wafer bonded so as to seal said first and second trenches to form said first and second cavities;

first and second resistor patterns formed on said second wafer so as to be within said first and second cavities, respectively;

a conductive film formed on said second wafer aligned with said third trench so as to form a capacitor with the conductive film on the bottom of said third trench;

fluid sealed in said first and second cavities that, when heated by current flowing through said first and second resistor patterns, raises the vapor pressure in said first and second cavities so as to flex said first and second diaphragms;

a third wafer bonded to said semiconductor wafer and having first and second valve seats formed thereon adjacent to said first and second diaphragms, and having first and second input channels formed thereon and coupled, respectively, to said first and second valve seats thereby forming first and second fluid passageways from first and second input ports, and having an output channel formed thereon so as to be coupled to both said first and second valve seats so as to form a fluid output passage for each of said first and second integrated valves.

Description Submit all comments and votes

Many industrial machines and industrial or manufacturing facilities are pneumatically powered. Pneumatic power provides very efficient actuation of machines, and is frequently used in robot machines for assembly line work. These types of machines are frequently controlled by computers or other logic circuitry. The logic circuitry decides the sequence of events that needs to occur, and generates electrical signals to cause same to occur as planned. When the sequence of events involves physical movement of portions of the machines which are driven pneumatically, there arises a need for a valve or conversion device which can convert the electrical control signals from the control logic into pneumatic control signals to drive the machine parts.

Since such machines often use many moving parts which are controlled by numerous individual pneumatic lines, it is frequently found that many such electric-to-fluidic valves are necessary. In such environments, the electric-to-fluidic valves need to be cheap, reliable, power efficient, small, and compatible with electronic interface circuitry between the valve and the computer or control logic.

In very precise robotic movement applications or other applications where very precise movement control is necessary, it is necessary to have precise control of the shape of the pneumatic control drive pulses. In other applications, such as gas chromatography, the shape of the fluid pulses entering the column must be precisely controlled to get precision assay data from the column. In either of these types of applications, the valves used to control the fluid flow must be precision valves which have little or no dead volume. Dead volume is the unknown volume which is trapped in a valve when it makes a transition from open to closed. This trapped fluid may escape into the stream thereby causing the shape of the fluid pulse to be altered from the desired shape. For example, in typical ga chromatograph systems, if a valve is used which has dead volume, the edges of the output fluid pulse entering the separation column (in terms of the volume of gas flowing at any particular instant in time) may not be vertical or sharply defined. Likewise, for precise robotic movement, it is desirable to have very sharp cut-offs for the pneumatic pulses used to drive robot fingers and arms to get precise positional control for the movement.

One known way of controlling the flow of a fluid using an electrical pulse is the electric-to-fluidic valve developed by Steve Terry of Stanford University. This valve uses a substrate such as silicon which has a thin membrane machined therein. This cavity is formed by the etching a hole almost completely through the substrate. This leaves a thin bottom wall for the cavity which is used as a flexible membrane. Attached to the side of the first substrate in which the membrane is formed is a second substrate which has a manifold type cavity etched therein with a passageway or nozzle formed in a wall of the manifold cavity for entering or escaping gas. The manifold cavity also has other ports formed therein to complete a fluid path into the manifold and out the nozzle or vice versa. The manifold cavity in the second substrate is positioned over the membrane of the first substrate such that when the manifold of the first substrate is flexed, it contacts a sealing ring formed around the nozzle of the manifold cavity thereby closing off the fluid flow path between the nozzle and the other ports into the manifold cavity. With the membrane of the first substrate in an unflexed position, the nozzle in the manifold cavity would not be pinched off, and fluid would be free to flow through the input port and the manifold cavity and out through the nozzle or vice versa. The membrane of the first substrate is forced to flex by mechanical forces exerted thereon by a piston. This piston is driven by a solenoid or other type of electromagnetic device.

One disadvantage of the above described valve configuration is that the solenoid requires a high power source, and is a large power consumer. Further, the solenoid or other electromagnetic device is large and heavy. The cavities in the first and second substrates could be formed with much smaller dimensions if it were not for the fact that the solenoid is large. Because the first and second substrates are silicon wafers which are etched using conventional planar photolithography techniques, it would be possible to make the electric-to-fluidic valve much smaller in dimension were it not for the solenoid. Such a prior art electric-to-fluidic valve construction is inefficient in its use of space. Because the solenoid is mechanically attached to the first substrate such that the piston of the solenoid pushes against the membrane in the first substrate and because the solenoid is large enough to consume much of the wafer space, generally only three such valve structures can be formed on a single silicon wafer. Such a structure is relatively expensive to build, and the bond between the solenoid and the glass is difficult to make. Generally, the solenoid is attached to a thick pyrex wafer by nuts and bolts. This form of attachment is both expensive to fabricate and a major source of failures. Further, such a structure has a moving part which can be another source of failure. The principal defect of such a structure, however, is the fact that the entire structure cannot easily be mass produced with planar lithography techniques. This is because the solenoid can not be manufactured by such techniques.

Another system which has been used in the past in the field of ink jet printing uses a principle used in the invention involving the tendency of fluids and gases to expand and to create higher pressures in a cavity when heated. The particular system which embodies this principle is a Hewlett Packard ink jet printer. This printer structure uses a print head which has a small cavity formed in or over a substrate. The substrate has formed thereon a resistive element, and the cavity is located over the resistive element. The cavity has a small ink jet nozzle therein through which ink may escape in small droplets when the pressure of ink in the cavity rises above the atmospheric pressure. In operation, such a structure will shoot out an ink drop each time a heating pulse is applied to the resistive element. The heat from the resistive element raises the temperature of the ink in the cavity thereby causing its vapor pressure to increase according to the laws of thermodynamics. When the pressure of the ink inside the cavity rises, one or more ink droplets are forced out of the cavity through the ink jet port in the cavity wall. Such a structure is an example of an unrelated application of a principle of thermodynamics which is used in the invention. As far as the applicant knows, no such application of the principle of expansion of a fluid in a confined cavity with increasing temperature has ever been used to control a fluid valve.

Thus a need has arisen for an electric-to-fluidic valve which may be mass produced cheaply using conventional planar lithography techniques, which does not use large amounts of energy, which is small and efficiently uses wafer space, which has no moving parts which slide across each other, which has sharp cutoff characteristics with little or no dead volume and which is compatible with the formation of interface or driver circuitry on the same silicon wafer in which the electric-to-fluidic valve is formed.

SUMMARY OF THE INVENTION

According to the teachings of the invention, there is provided an electric-to-fluidic valve utilizing the principle of expansion and pressure rise of a fixed volume of gas or fluid when heated to deflect a flexible wall or thin membrane forming one or more walls of the cavity in which the gas or fluid is contained. The deflection of the membrane may be used to seal or unseal a fluid passageway from an input port, through a manifold cavity and out an output nozzle or vice versa. The valve may also be operated linearly to provide a linear range of fluid control, i.e., the valve may be controlled to modulate the rate of fluid flow through the valve in accordance with the magnitude of a control signal. The deflection of the membrane may also be used as a sensor indication for purposes of determining temperature changes or the magnitude of other phenomena to be measured.

The structure of an electric-to-fluidic valve according to the teaching of the invention may have numerous variations. However, the principal elements of each design will include a cavity formed in a substrate where one wall of the cavity is a thin, flexible membrane. The cavity encloses a fixed number of moles of gas or fluid, and there will be some method or means of raising the temperature of the fluid in the cavity so as to cause the vapor pressure to rise in the case of a fluid or to cause expansion and increased pressure in the case of a gas. This heating of the material in the cavity may be accomplished in any one of a number of ways. One way is the use of a resistive heating element on one wall of (in the case of diffused resistors, in one or more walls) or located somewhere inside the volume of the cavity such that electrical current may be passed through the resistive element to generate heat and heat the fluid trapped in the cavity. Other possible methods of heating the fluid in the cavity include radio frequency heating of the material in the cavity by beaming radio frequency energy into the cavity from an energy source located outside the cavity, or through conductive, convective or radiated heating of the material in the cavity. Further, optical heaters could be used to provide an optical to fluidic conversion. In such a structure light can be beamed into the cavity having walls of transparent material. This light either heats up the vessel walls (one or more walls may be coated with a material which absorbs the light energy) or heats up the material in the cavity directly if the material in the cavity is a dark fluid or gas which can efficiently absorb the light. The light energy can be transmitted from the source either by radiation or with the aid of a light pipe or a fiber optic light guide. Such an optical to fluidic conversion provides reliable pneumatic control in an electrically noisy environment since electrical noise is not picked up by radiated light beams or by fiber optic light pipes.

In other embodiments, it is possible to change the temperature of the fluid in the chamber having the thin membrane as one wall (hereafter called the membrane chamber) by cooling the fluid trapped in the chamber. This may be done using Thompson or Peltier coolers. Other types of cooling mechanisms may also be employed such as simple refrigeration systems, or radiative, conductive or convective coolers. According to the teachings of the invention, any method of controlling the temperature of the fluid in the membrane chamber will suffice for purposes of practicing the invention.

Although the structure of devices built in accordance with the teachings of the invention will vary drastically from one application to another as can be surmised from the above discussion of the different types of heating structures which may be used, a typical structure utilizes a silicon-pyrex sandwich for the membrane cavity and heating structure. The membrane chamber is formed in a silicon wafer by etching a trench in the wafer substantially completely through the silicon wafer but stopping short of the opposite side of the wafer by a margin which is equal to the desired thickness of the membrane of the membrane chamber. Other signal processing circuitry, such as power transistors or full feedback control systems with multiplexed input and output ports, may have been previously fabricated on the balance of the wafer in conventional processing. This circuitry can be used in conjunction with the electric-to-fluidic valve formed by the membrane chamber thereby forming a valve or transducer with its own interface circuitry located on the same silicon wafer as the valve itself using compatible processing steps. The same of course is true for sensor applications where the membrane chamber is used as a transducer. The signal processing or other circuitry built elsewhere on the wafer may then be used to signal process, condition or otherwise deal with the output signal from the transducer for their operation.

The surface of the silicon substrate having the membrane as a part thereof is sandwiched with another wafer in which a manifold having an input port and a nozzle is formed (the nozzle may be the input port and the other port may be the output port also). The fluid manifold position is keyed so that when the second wafer is attached to the first wafer, the nozzle and a sealing ring around same are located within the path traveled by the membrane during deflection. Deflections of the membrane change the cross-sectional area of the fluid communication path between the input port and the output port of the fluid manifold. If the deflection is large enough, the membrane seats completely on the sealing ring around the nozzle and completely cuts off flow through the nozzle.

The invention may be practiced as a conversion type valve or as a transducer. The term transducer as it is used herein should be construed as meaning a device which produces an output signal which characterizes the magnitude of some parameter which is being measured with the transducer. The parameter being measured may be either the temperature of the fluid in the chamber or some other parameter which affects the temperature of the fluid in the chamber. Examples of the latter would be the magnitude of current flow through a resistive element in the chamber, the intensity of a light beam and so on. In other embodiments, the sensor structure according to the teachings of the invention could measure pressure by having the pressure to be measured applied to the membrane. The changes in pressure of the fluid in the chamber in response to deflections of the membrane under the influence of the force or pressure to be measured are translated into transient changes in temperature of fluid in the membrane chamber in accordance with the laws of thermodynamics. These changes in temperature can be sensed with a thermocouple or other temperature sensing device to generate the output signal which characterizes the pressure or force to be measured.

In accordance with the teachings of the invention, a temperature sensor of the ambient temperature can be fabricated by fabrication of an electric-to-fluidic valve without any external or internal means of changing the temperature of the fluid in the membrane chamber. A flow meter can then be placed in the fluid communication channel controlled by the valve membrane such that the flow rate through the valve can be measured as a function of the ambient temperature affecting the fluid in the membrane chamber. That is, changes in ambient temperature will cause changes of the temperature of the fluid in the membrane chamber. These temperature changes will be translated to deflection changes in the membrane of the membrane chamber thereby modulating the fluid flow through the fluid communication channel of the valve. This changing flow rate could be amplified using known techniques such as staged pneumatics and used directly in a pneumatically or hydraulically driven control system.

There is also disclosed herein an integrated pressure regulator and an integrated flow regulator. The integrated pressure regulator uses two integrated valves of the type disclosed herein which serve to control the pressure in an output port situated between the valves. The output port pressure is sensed by a capacitive transducer where one plate of the capacitor is fixed on the substrate and the other is formed on a flexible diaphragm which forms one wall of an evacuated chamber formed in the substrate. Changes in pressure cause the diaphragm to flex and cause the second capacitive plate to move closer to or further away from the first plate. These changes in capacitance are used to determine the pressure in the output port and to generate an error signal. The error signal is used to generate control signals which control the two valves. These valves are integrated on the same die with the output port and the capacitive sensor. One valve couples a high pressure source to the output port while the other valve couples the output port to a low pressure effluent sink.

The integrated flow regulator utilizes one integrated valve according to the teachings of the invention and a flow channel. Three resistor elements are placed along the flow channel. The middle resistor is driven with a current that causes this resistor to have a constant temperature. The heat caused by the middle resistor diffuses toward the cooler resistor temperature sensors on either side of the middle resistor. But because of the flow of material in the flow channel, less heat reaches one resistor temperature sensor than reaches the other and this causes a temperature differential. The amount of this temperature differential is related to the flow rate and is sensed by control circuitry coupled to the resistor temperature sensors. The flow rate so determined is compared to a desired flow rate to generate an error signal by conventional circuitry. This error signal is used to control the valve such that the flow rate is altered toward the desired flow rate.

A better understanding of the different types of structures which can be made in accordance with the teachings of the invention and the different methods of making these structures may be had by reference to the drawings herein for which a brief description of the drawings follows.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a cross-sectional view of an electric-to-fluidic valve of the invention.

FIG. 2 is a diagram of one porting arrangement that may be used with the valve of the invention.

FIG. 3 is a diagram of another porting arrangement that may be used.

FIG. 4 is a diagram of another porting arrangement that may be used.

FIGS. 5 through 14 are cross sectional views of wafers 1 and 2 at various stages in one process sequence used to make the valve of FIG. 1.

FIG. 15 is a cross sectional view of a polyimide membrane valve structure.

FIGS. 16 through 27 are cross sectional views of the processing of various wafers necessary to form the polyimide membrane valve of FIG. 15.

FIGS. 28A and 28B are plan views of two different configurations of the polyimide membrane valve of FIG. 15 showing alternative arrangements for the configuration of the membrane relative to the channel size.

FIGS. 29 through 39 are process sequence drawings for the process to make another polyimide membrane valve.

FIGS. 40 through 48 are process sequence drawings for the process to make another polyimide membrane valve with a "folded" polyimide membrane.

FIGS. 49 through 52 are process sequence drawings illustrating the process steps of the preferred form of encapsulation of material in any of the embodiments using electroplating on an etched resistor pattern on the wafer which seals the membrane cavity.

FIG. 53 is a cross sectional view of a solid state heat pump valve embodiment.

FIG. 54 is a view of a micropositioning embodiment of the invention where the excursion of the membrane is used to position an object.

FIG. 55 is a diagram of a robotic embodiment using tactile feedback.

FIGS. 56 through 64 are cross sectional views through the structure of the tactile feedback transducer at various stages in the process of manufacture.

FIG. 65 is a cross sectional view of the tactile transducer in the finished state with the bubble inflated.

FIG. 66 is a diagram of the row and column addressing system of an array of tactile transducers.

FIGS. 67 through 77 are cross sectional drawings of the structure of a second embodiment of a tactile actuator at various stages in the process of its manufacture.

FIG. 78 is a cross sectional diagram of a variable focal length lens.

FIG. 79 and 80 are cross sectional diagrams of eutectic bonds between silicon wafers and wafers of another type of semiconductor.

FIG. 81 is a cross sectional view of the preferred embodiment of a valve according to the teachings of the invention.

FIG. 82 is a plan view of the mesa and channel structure formed in the lower wafer of the valve shown in FIG. 81.

FIG. 83 is an alternative mesa structure with higher off resistance than the structure of FIG. 82.

FIG. 84 is a cross sectional view of an integrated pressure regulator using integrated valves.

FIG. 85 is a cross sectional view of an integrated flow regulator according to the teachings of the invention.

FIG. 86 is a cross sectional view of an alternative form of flow regulator.

FIG. 87 is a cross sectional drawing of the nitride beam encasing the resistor element which may be used in any of the embodiments disclosed herein.

FIG. 88 is a plan view of the resistors of FIG. 86.

FIG. 89 is a cross sectional view of a bistable embodiment of the valve of FIG. 81 or any other valve embodiment.

FIG. 90 is a cross sectional view of another bistable embodiment of the valve of FIG. 81 or any other valve embodiment.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT

Referring to FIG. 1, there is shown a cross-sectional view of the preferred embodiment of an electric-to-fluidic valve in accordance with the teachings of the invention. The valve is comprised of a membrane chamber 10 which is etched into a substrate 12 of [100] orientation silicon in the preferred embodiment. The basic structure of the invention may be realized with many different type of materials including electroformed steel, plastics and many others. However, in the preferred embodiment shown in FIG. 1, a three layer sandwich consisting of silicon wafer 12, another silicon wafer 30 and a pyrex wafer 22. Wafer 22 is made of Pyrex 7740 in the preferred embodiment. Silicon was chosen for the other wafers because it can serve as a substrate for the formation of electronic circuitry for signal processing or interface purposes. It is likely, that the signal processing or interface circuitry will be placed on another chip and connected in hybrid fashion using known techniques. However, it is possible to build other electronic circuitry on the same wafer as the valve. The decision on which method to use may be made by the user on other criteria than whether making circuitry on the same wafer as the valve or sensor is possible.

The membrane chamber 10 is defined by six walls, four of which are silicon [111] planes and two of which, walls 14 and 16, are shown in cross section. Because the membrane chamber 10 is anisotropically etched with KOH etchant which will not etch the [111] orientation silicon plane, the walls 14 and 16 form angles of 54.73 with the surface of the membrane 18. The remaining wall of the membrane chamber 10 is the thin, flexible membrane 18.

The volume of the membrane chamber 10 is fixed except when the membrane 18 flexes. A fixed quantity of a gas or fluid is sealed in the membrane chamber 10. This may be done during the sealing of the membrane chamber by bonding a pyrex wafer 22 to the top surface of the wafer, i.e., the surface of the wafer 12 having the most positive z coordinates and normal to the z axis. The pyrex wafer has formed thereon a photolithographically etched resistor pattern 20 in the preferred embodiment which serves to allow the contents of the membrane chamber 10 to be electrically heated. Pyrex was chosen for the wafer 22 because it is transparent to light and would allow optical energy to be beamed into the membrane cavity 10 through the pyrex to heat the trapped material therein. Also, pyrex forms a hermetic seal with silicon at a relatively low temperature (300.degree. C.). Such an optical to fluidic embodiment is symbolized by the presence of the light pipe 19 which serves to guide light energy to the cavity 10 and direct it into the cavity. The light pipe 19 may be any light guide such as fiber optic cable. This would allow an optical to fluidic valve to be constructed in some embodiments where the resistive heater 20 is inappropriate or is not convenient.

Further, the light pipe 19 only symbolizes the many different forms the heater element may take. For example, no heating element at all may be used and the temperature of the ambient environment may be used to cause the heating and cooling of the material in the chamber. Conduction or convection heating might also be used by heating the wafer 22 and allowing this wafer to conduct the heat to the cavity 10. The conduction heating would be by direct contact of the heat source with the wafer 22 and convection heating would be by flowing hot gas or liquid over the wafer 22. In these latter environments, the material of the wafer 22 should be selected so as to conduct heat well so as to not slow response to the heat signals. In the embodiment shown in FIG. 1, the pyrex wafer 22 serves as a sealing member for the membrane cavity and as a substrate upon which a resistive heater may be formed by planar photolithography. Contact to the resistive heater 20 may be made in any conventional way. One way is through formation of a hole in the glass. The resistive heater 20 is made of aluminum in the preferred embodiment, but a heater made of chromium and gold or any one of a number of materials will also work. In some embodiments, the resistive heater 20 will be formed with a thermocouple as a part thereof for coupling to external temperature measuring equipment. This will allow the temperature of the resistive element to be measured.

In other embodiments, the resistive coefficient of temperature may be used to monitor the temperature of the resistive element. That is, the temperature of the resistive element 20 may be known by measuring the amount of current flowing therein.

The material chosen for the heater element 20 must not react with the material trapped in the membrane chamber 10 either at low temperature or when heating the material. In the preferred embodiment, a thin protective coating (not shown) is formed over the heating element to prevent any reaction with the material being heated. This thin coating may be any material which will effectively seal the heating element while not degrading the heat conduction from the resistive element 20 to the material being heated substantially. In embodiments where there is no danger of reaction between the heating element and the material trapped in the membrane chamber, the coating may be omitted. The conductors of the resistive element may be brought out to bonding pads on the outside edges of the pyrex wafer.

The membrane chamber 10 is filled with a fluid or gas which is selected in accordance with a criteria which will described in more detail later. This filling may be done while the pyrex wafer 22 is being attached, or after attachment by means of forming a port into the membrane chamber 10, filling the chamber with gas or fluid, and then sealing the port. Generally speaking, the material selected to fill the membrane chamber 10 is picked based upon its activation energy such that at the highest ambient temperature likely to be experienced by the valve, the pressure of the fluid or gas in the membrane chamber will not be so great as to cause deflection of membrane 18 to pinch off fluid flow through the fluid manifold and nozzle. Also, the fluid chosen to fill the cavity so as to maximize the ratio: delta pressure/ delta energy input meaning the fluid is chosen to get the maximum change in pressure for a unit change in the energy input. Optimizing this ratio minimizes the power consumption which is very important in some applications. Also, the fluid must be chemically inert so as to not cause adverse reactions with the material of the valve and other materials with which it might come into contact. In the more general sense, the flexure of the membrane may be used to modulate the cross section of the fluid flow path through the manifold cavity 24.

The pyrex wafer 22 has a resistor element 20 formed thereon through which electrical current may be passed to heat up the contents of the membrane chamber. A second wafer 30 having a manifold chamber 24 formed therein is attached to the first wafer 12. The manifold has an input port (not shown) and a nozzle 32 formed therein. The nozzle 32 has a sealing ring 28 formed around its perimeter. When the contents of the membrane chamber 24 are heated, and the membrane 18 deflects to the position shown in dotted lines in FIG. 1, the membrane 18 seats on the sealing ring 28 thereby cutting off all pneumatic flow through the nozzle 32. The term pneumatic flow is used here to indicate either gas or liquid flow. As the material trapped in the membrane chamber 10 cools, it contracts thereby decreasing the pressure in the membrane chamber 10 and causing the membrane 18 to deflect away from the sealing ring 28 around the nozzle.

It is possible to get a reverse flexure of the membrane 18 into the membrane chamber 10 if too much pressure is present in the nozzle 32/manifold cavity 24. This will be referred to as an overpressure situation. If this occurs, the membrane can break if the deflection becomes too great. One way is to make the depth, d in FIG. 1, of the membrane cavity small enough such that when overpressure occurs, the surface 21 of the membrane cavity stops the deflection of the membrane 18 before the breaking point or elastic limit is exceeded. A second method of preventing destruction of the membrane 18 is to fill the membrane cavity with liquid. Because of the incompressible nature of liquid, as the volume of the membrane cavity 10 is decreased with the reverse flexure of the membrane 18 when overpressure occurs, the pressure in the membrane cavity rises rapidly. This rise in pressure counteracts the flexure and prevents the membrane from flexing past the breaking point.

The anisotropic etch step to form the membrane chamber 10 is chosen so as to give precise control of the dimension A of the membrane as marked in FIG. 1. It is important to have precise control over the dimensions of the membrane chamber and the lateral positions of the walls 14 and 16. This i