Patent Application
20140224019
The present disclosure generally relates to pressure sensors. More specifically, but not exclusively, the present disclosure is concerned with piezoelectric pressure sensors and with an array of piezoelectric pressure sensors for wave parameters measurement, and with methods for producing piezoelectric pressure sensors.
The advance in microfabrication techniques makes it possible to produce very accurate and versatile sensors used in a variety of systems involving mechanical, electrical, optical and biological sensing. A major group of mechanical sensing devices use the piezoelectric effect to obtain high speed detection of a mechanical displacement, a pressure or a force. Because of their fast responses, these types of sensors are very popular in applications involving wave propagation, either shock waves or acoustic waves. Using the proper sensors and signal conditioning systems, it is possible to provide an accurate history of pressure variation in a certain location, resulted from passage of the waves. However, to obtain a complete picture of wave propagation in the medium it is necessary to probe the medium in different locations, so that the direction and the speed of the wave could be measured, in addition to its amplitude time history.
There are many applications for simple devices that could measure the local velocity vector of a mechanical wave. Shock tubes have been known in the art of fluid mechanics for quite some time. Shock tubes may be used in the study of unsteady high speed flows. To acquire practical information on the speed and propagation of a wave in a shock tube, a number of sensors are typically installed along the length of the shock tube in such a manner as to detect change in at least one physical property of a gas contained in that shock tube.
Sensing the speed of a wave may, in theory, be made using, for example, two pressure transducers installed along a shock tube. Measuring the time taken by the wave to travel between the two transducers and knowing the distance between them allows for the computation of the average wave speed over this distance. The wave velocity may have fluctuated when travelling from one transducer to the next, therefore such a setup allows for measuring the average speed.
Furthermore, measuring the direction of propagation of a wave may, in theory, be made using more than two sensors, wherein this plurality of sensors is not located on a straight line. However, such a simple setup may render the measurements inaccurate. This is because the speed and direction of a pressure wave jointly define a velocity vector whose properties may depend on the position of the wave. To obtain an accurate measurement of the local wave velocity vector therefore requires the plurality of sensors to be in close proximity. This is difficult to achieve with current commercial pressure sensors which are packaged individually and which each occupies a fairly large surface of many square millimeters.
The same situation may take place in components of turbomachines, such as fans, compressors and turbines. Many flow phenomena in gas turbines are unsteady, meaning that the flow properties vary in time at a certain fixed location, leading to wave propagating in various directions. For example, some or all blades within a compressor may stall and the pressure at a given location may vary in time. To identify the amplitude, speed and direction of stall waves in such a situation would require the use of many pressure sensors in close proximity, a configuration difficult to achieve in practice due to the relatively large size of actual pressure sensors and the limited space available in typical turbomachines.
Another situation takes place in microdevices where the space available for measuring the speed of waves is severely limited. Microscale shock tubes have been introduced for this purpose. Such shock tubes may have cross sections of the order of a few micrometers. Obviously, measurements of physical properties of gas taking place in such small scales cause important difficulties, and the size of sensors cannot exceed the size of the channel of the microscale shock tube. Moreover, in operation, these sensors need to be put in direct contact with the flow of gas and have a reaction time sufficiently fast to detect gaseous pressure changes occurring at nanosecond scales.
Conventional pressure sensors require the presence of a mechanical element, such as a membrane, having a shape that may be altered under pressure in a manner that the shape alteration may be measured. Miniaturization of the sensors implies a very small and very thin membrane, difficult to fabricate, whose shape alteration that may only be measured using technologically complex methods, such as with an atomic force microscope, for example.
There therefore exists a need for a method for fabricating sensors that are simple to operate and yet are sufficiently small that an array of them may be packaged in a small area and volume.
According to the present disclosure, there is provided a method for manufacturing a piezoelectric sensor. An electrical barrier is formed on top of a silicon substrate. A bottom electrode layer defining a bottom positive electrode section and a bottom negative electrode section is deposited on top of the electrical barrier. A piezoelectric layer is deposited on top of the bottom electrode layer. A positive electrode connection area and a negative electrode connection area are etched, through the piezoelectric layer. A top electrode layer is deposited on top of the piezoelectric layer. The top electrode layer is making contact with the bottom electrode layer through the positive and negative electrode connection areas and defines a upper positive electrode section and a upper negative electrode section. A sensing area is created, in the piezoelectric layer, in an area of overlap between the upper positive electrode section and the bottom negative electrode section or between the upper negative electrode section and the bottom positive electrode section.
According to the present disclosure, there is also provided a piezoelectric sensor comprising a silicon substrate, an electrical barrier on top of the silicon substrate, a bottom electrode layer on top of the electrical barrier, a piezoelectric layer on top of the bottom electrode layer, and a top electrode layer on top of the piezoelectric layer. The bottom electrode layer defines a bottom positive electrode section and a bottom negative electrode section. The piezoelectric layer defines a positive electrode connection area and a negative electrode connection area. The top electrode layer makes contact with the bottom electrode layer through the positive and negative electrode connection areas and defines a upper positive electrode section and a upper negative electrode section. A sensing area is defined, in the piezoelectric layer, in an area of overlap between the upper positive electrode section and the bottom negative electrode section or between the upper negative electrode section and the bottom positive electrode section.
According to the present disclosure, there is also provided a method of measuring an amplitude, a speed and a direction of propagation of a shock wave in a shock tube. A piezoelectric sensor array comprising a plurality of piezoelectric sensors disposed in a pre-defined configuration is attached to the shock tube. The piezoelectric sensor array is connected to a signal analysis device. The shock wave is initiated in the shock tube. The signal analysis device detects an arrival time of the shock wave at each of the plurality of piezoelectric sensors.
The present disclosure further relates to a smart pressure sensor array comprising a plurality of sensors packaged in close proximity in the sensor array, and one or more wired connections for connecting the sensors to a data acquisition system. The sensor array provides the data acquisition system with pressure time histories at an individual location of each sensor of the array.
In the appended drawings:
The foregoing and other objects, advantages and features of the present disclosure will become more apparent upon reading of the following non-restrictive description of illustrative embodiments thereof, given by way of example only with reference to the accompanying drawings.
Generally stated, sensors and sensor arrays described herein may be applied to measuring a physical property of a fluid such as a gas or liquid, for example pressure of the gas in a shock tube. Many pressure sensors may be fabricated and packaged into a device comprising an array of sensors occupying no more space than a single conventional pressure sensor. A possible application of the arrays of theses sensors with particular geometries comprises measuring the amplitude, speed and the direction of propagation of waves in a fluid. Simultaneous local measurement of wave amplitude, speed and direction with great spatial and temporal resolutions may be obtained with high accuracy. Another possible application is the measurement of wave speed and wave propagation direction in turbomachinery such as fans, compressors and turbines. A non-limitative example of implementation is an array of sensors comprising five (5) to eight (8) sensors, the sensors being positioned in a pre-defined configuration or geometry, for example a circular geometry or a cross-shaped configuration.
Piezoelectric direct sensing pressure sensor arrays suitable for various applications have been fabricated and tested. The sensor arrays exhibit small size, for each sensor, of the order of a few microns, and fast time response, with a natural frequency which may exceed 1 GHz. Fabrication of such piezoelectric sensors involves in part processing of piezoelectric material such as, for example, Lead Zirconate Titanate (PZT) thin films.
A circular configuration of the sensor arrays provides a good resolution in the measurement of the direction of propagation of the wave. A circular array of eight (8) sensors is sufficient to obtain a small deviation between theoretical expectations and actual laboratory results. For very high speed applications, a simpler, cross-shaped array actually requires less post-processing calculation power.
While the present disclosure relates mainly to applications of the sensor array to large scale and microscale shock tubes as well as turbomachines, those of ordinary skill in the art will appreciate that the sensor array may also be used in many other applications where small sensing devices are used.
Turning now to drawings,
A smart pressure sensor array 102 uses a plurality of sensors Si packaged in the same device 102 in close proximity to provide the pressure time histories at each individual sensor location. By simultaneously analyzing the pressure time histories at different positions it is possible to very accurately calculate the speed and also the direction of propagation of the wave 104. The sensors Si may be wired individually to a data acquisition and analysis system 106 or the signals from different sensors Si of the array 102 may be coded at an encoder 108 and merged into one signal that may be transmitted with a single wire 110 and then separated by a decoder 112 and supplied to the data acquisition and analysis system 106, as shown in
h
i
=h
c
−R cos(φi−θ) (1)
Assuming a constant local speed μs of the shock wave 302, dividing equation (1) by the speed of the shock wave, an arrival time of the shock wave 302 on each sensor Si may be calculated using equation (2):
In equation (2), tc and ti designate times of arrival of the shock wave 302 at the center 308 of the circular sensor array 300 and at each sensor Si, respectively. Equation (2) is of particular interest in that it relates the time of arrival ti of the shock wave on each sensor (which is a measurand) and allows postprocessing calculations for obtaining the speed and the direction of propagation of the shock wave.
Therefore, to find the speed and the direction of propagation of the shock wave, a cosine curve may be fitted to the data as illustrated in
Some other geometries, for example a cross shape array of 5 sensors, are also possible.
t
2
=t
c−(R/μs)sin θ (3)
where tc and t2 are the time of arrival of the shock wave or mechanical wave to the center of the device and the outer sensor S2, respectively. Equation 3 again relates the arrival time of the shock wave on each sensor (which is a measurand) to the speed and the direction of the shock wave. Doing the same calculation for all the outer sensors S1, S2, S3 and S4 we have:
t
c
−t
1=(R/μs)cos θ (4a)
t
c
−t
2=(R/μs)sin θ (4b)
t
c
−t
3=−(R/μs)cos θ (4c)
t
c
−t
4=−(R/μs)sin θ (4d)
and therefore the wave direction θ may be obtained from:
and the wave speed obtained from:
Since the equations (5) and (6) are over-determined and there are four (4) different equations for each unknown θ and μs, these two sets of equations may be used to obtain the final result by averaging over the computed values or used to eliminate spurious or faulty measurements.
As a non-limitative example, the circular sensor array 300 of
An example of a method of fabrication for the sensor Si delimited by the lines A-A and B-B of
The cross-sectional elevation views of
A first operation 530 comprises a thermal oxidation of the polished face of the substrate 502. This operation 530 produces an approximately 600 nanometers (nm) thick oxide layer 504 acting as an electrical barrier on top of which other layers will subsequently be added.
At operation 540, the oxide layer is etched away, using any suitable etching process known to those of ordinary skill in the art, for example Inductively Coupled Plasma (ICP) etching with CF4 chemistry, in regions 506 and 507 (corresponding to lines A-A and B-B of
A bottom electrode layer comprising a bottom ground electrode section 508 and a bottom live electrode section 509 of the sensor Si is produced in operation 550 by depositing a 15 nm thick sub-layer of titanium forming an adhesion layer on the oxide layer 504 and, then, a 150 nm thick sub-layer of platinum as bottom electrodes. Both platinum and titanium sub-layers may be deposited in an electron beam evaporator or a sputtering chamber and annealed at 570° C. in nitrogen ambient. As can be seen in
At operation 560, a piezoelectric layer 510, for example a PZT layer, having for example a 50 to 500 nm thickness, is deposited by the sol-gel method on the bottom electrode sections 508 and 509. The sol-gel method is a wet-chemical technique starting from a chemical solution (or sol) which acts as a precursor for an integrated network (or gel) of either discrete particles or network polymers, as described in more detail at http://en.wikipedia.orq/wiki/Sol-qel. The operation 560 may include a number of cyclic depositions, pyrolyzing and annealing operations to obtain a desired thickness of the piezoelectric layer 510. In this manner, for example, a good quality sol-gel derived Lead Zirconate Titanate (PZT) layer can be developed, free of cracks, by overcoming problems such as diffusion and oxidation of titanium and residual stresses in the platinum sub-layer. Of course, any other material capable of producing an electrical field as a result of compression may suitably replace PZT.
The sol-gel derived PZT layer 510 features an extremely large dielectric constant (in a range of 800-1100), an increased piezoelectric response and poling efficiency. To electrically connect the bottom electrode layer to a top electrode layer, which will be added later as described hereinafter, the PZT layer 510 is etched at circular areas 512 (see
In a next operation 570, a top electrode layer comprising a top ground electrode section 514 and a top live electrode section 515 are produced by depositing a 15 nm thick sub-layer of titanium forming an adhesion layer on the PZT layer 510 and, then, a 150 nm thick sub-layer of platinum forming the top electrodes. During operation 570, the top electrode sections 514 and 515 connect with the bottom electrode sections 508 and 509, respectively through the etched areas 512 in the PZT layer 510. The same methods of deposition as employed for the bottom electrode section 508 and 509 may be used. As can be seen in
The overlapping geometry of the bottom electrode layer, piezoelectric layer and top electrode layer described hereinabove allows to easily create patterns on the various layers to define a sensor having an active area 513 and deactivate the rest of area on the surface of the substrate by shorting it without removing the piezoelectric material from the deactivated area on the substrate that may cause the delamination of the platinum bottom layer. Those of ordinary skill in the art will appreciate that operations of
Then, at operation 580, the silicon substrate 502 may be etched using, for example DRIE, throughout at the regions 506 and 507 to extract a ring-shape chip comprising eight (8) sensors from the substrate. Finally, at operation 590, wires such as fine gold wires 518 and 519 are soldered at areas 512 to respective electrodes formed by electrode sections 508 and 514 and electrode sections 509 and 515.
The foregoing description refers to elements 508 and 514 as bottom and top ‘ground’ electrode sections, respectively, and to elements 509 and 515 as bottom and top ‘live’ electrode sections, respectively. In other realizations, elements 508 and 514 may form a live electrode while elements 509 and 515 may form a ground electrode. More generally, any connected pair of bottom and top electrode sections may act as a positive electrode, the other pair of bottom and top electrode sections acting as a negative electrode. It is understood that the terms ‘positive’ and ‘negative’ reflect relative voltages between complementary pairs of electrode sections.
The single ring-shaped array 300 comprising eight (8) piezoelectric sensors Si of
In operation, when installed in a sensed device such as a shock tube, each of the eight (8) piezoelectric pressure sensors reacts to a shock wave or mechanical wave pressure applied to the PZT layer 510 to produce an electric signal through the electrode formed by electrode sections 508 and 514 (and wire 518) and the electrode formed by electrode sections 509 and 515 (and wire 519). Electric signals obtained from the sensors may be amplified and are supplied to a signal analysis device. Signal analysis is based on a mathematical model, which may for example be based on Equations (1) and (2) when the pre-defined configuration of the sensor array is circular as shown for example in
A challenge in the microfabrication of these sensors is the elimination of the fine gold wires 518 and 519 from the design of
Improving on this design involves wiring out the thin film piezoelectric structure on the front of substrate to the back side of the substrate while keeping the smooth and sealed surface of the sensors. A challenge in the microfabrication is thus to integrate the sensors' structures with electrical vias on a substrate. In fact there are some known approaches to create vias in silicon substrates. However, these approaches cannot be integrated with the thin film piezoelectric development procedure described above and a new approach for the compatible microfabrication of vias is presented.
Therefore another example of a method of fabrication for the sensor Si delimited by the lines A-A and B-B of
Operation 915 starts with production of a Silicon On Insulator (SOI) substrate. The substrate has a 30 μm thick <100> silicon device layer 902, 2 μm thick Buried Oxide (BOX) layer 904 forming an upper electrical barrier, and 300 μm thick lower handle layer 906. These thicknesses are chosen to meet criteria such as mechanical strength, ease of silicon dry and wet etch that will be performed later on, enhanced electrical insulation and less capacitive parasites.
At operation 920, to make a reliable mask for the silicon wet etch, 7 μm thick Plasma Enhanced Chemical Vapor Deposited (PECVD) oxide layers 908 and 912 are respectively deposited on a front side and on a back side of the substrate. This will allow the substrate to withstand the long 300 μm silicon wet etch in potassium hydroxide (KOH) solution.
At operation 925 the oxide 912 on the lower handle layer 906 is etched 914 using Advanced Oxide Etching (AOE) and an ordinary photoresist mask to pattern the oxide mask. Proper care in executing this lithography process will prevent growing of any small flaw during wet etch, which could make the substrate unusable for the next steps.
During operation 930 the lower handle layer 906 is etched through to arrive at the BOX layer 904. The anisotropic etch of silicon in the KOH solution results in the formation of pits 916 in the lower handle layer 906 with inclined and smooth walls.
At operation 935, while protecting the BOX layer 904, the mask oxide layers 908 and 912 are on both sides are wet etched in hydrofluoric acid bath. To protect the BOX layer 904 in the bottom of the pits 916, the photoresist is spin coated on the lower handle layer 906 and wiped on the top surface, followed by plasma burning of the residues of photoresist. This is repeated for a few times until the bottoms of all the pits 916 are protected. Then the PECVD masks are removed in, for example hydrofluoric acid or any other suitable etching solution.
At operation 940, isolated islands 922 are created on the device layer 902 by Deep Reactive Ion Etching (DRIE) of annular trenches 924 down to the BOX layer 904. Since the surface of the sensors should not include any topography, these trenches should be closed with dielectric material. Therefore, the trenches should be as narrow as possible.
At operation 945 the trenches 924 are covered by deposition of 4 μm thick PECVD oxide 926 on the device layer 902. If the trenches are not completely covered, the processing materials in the next steps may enter into the trenches 924 and may short the isolated islands.
At operation 950, the PECVD oxide 926 is removed from the surface, using AOE, except at annular areas 928 over the trenches 924. The PECVD oxide 926 will later be replaced by thermal oxide except over the trenches 924.
At operation 955, the entire substrate is thermally oxidized to 1.5 μm oxide thickness after RCA (Radio Company of America) cleaning. This creates thermal oxide layers 932 and 933 by oxidizing both surfaces of the substrate as well as inclined walls 934 of the pits 916. This terminal oxide layer 933 becomes a lower electrical barrier. This process also increases the thickness of the BOX layer 904 in the pits 916.
At operation 960, the thermal oxide layer 932 is etched on isolated islands 936 using AOE to electrically reach to the device layer 902 from the front side of the substrate.
At operation 965, to electrically reach to the device layer 902 from the back side of the substrate, the BOX layer 904 is etched at annular areas 938, using AOE. Spray coated photolithography is used to deposit a uniform layer of photoresist (not shown) on the non-planar surface of the lower handle layer 906.
Bottom electrode sections of the sensors are realized by deposition of a layer 942 comprising 150 nm of platinum and 15 nm of titanium as an adhesion layer, at operation 970. This operation also includes the same metal deposition on the back side of the substrate, forming layer 944. The platinum layers 942 and 944 are deposited in an electron beam evaporator and annealed at 570° C. in nitrogen ambient.
At operation 975, a PZT layer 946 is deposited on the layer 942 by the sol-gel method. To electrically connect the top and bottom electrode sections, the PZT film is etched at the desired locations 948.
At operation 980, top electrode sections 952 are realized by deposition, on top of the PZT layer 946, of similar layers of platinum and titanium of the same thicknesses as in operation 970, these layers of platinum and titanium being etched at desired locations.
At operation 985, the device layer 902 is deep etched to the BOX layer 904 at annular area 954 followed by etching of the lower handle layer 906 to the BOX layer 904 are at annular area 956 to extract circular chips from the SOI substrate.
The chip 1002 will have a reliable support to be exposed to the fluid flow and to establish the electrical connections. Referring again to
A method for measuring the amplitude, speed and direction of propagation of a wave in a fluid flow is described below.
Measuring the speed and propagation direction of a wave in a fluid flow may be accomplished using a sensor array, for example the ring-shaped array 300 of eight (8) piezoelectric sensors Si of
Returning to the description of
Additionally to their capacity to transform a mechanical signal into electrical impulses as sensors or receivers, piezoelectric materials can also produce mechanical waves if electrically excited appropriately, as emitters. Accordingly, the sensors arranged and/or fabricated as described in the present disclosure may also individually be operated as emitters. Individually tailoring the wave signal simultaneously produced by each emitter in an emitter array, operating as a phased array, then allows control over the features of the mechanical wave beam produced by the emitter array. Furthermore, the sensors may successively be used as emitters, to produce for example an ultrasound beam of short duration, known as a pulse, in the medium into which they are in contact, and then shortly thereafter as receivers to analyze an echo produced by the reflection and refraction of this pulse with various inhomogeneities in the medium. Operating individual sensors and sensor arrays in this pulse-echo mode would, for example, allow for the detection of defects in the medium or imaging the various features in the medium, as it is already done in ultrasound non-destructive testing (NDT) and imaging. So the sensors and sensor arrays described here may be used for NDT and imaging, with new capabilities owing to their small size.
It is to be understood that the present disclosure is not limited in its application to the details of construction and parts illustrated in the accompanying drawings and described hereinabove. The present disclosure is capable of other embodiments and of being practiced in various ways. It is also to be understood that the phraseology or terminology used herein is for the purpose of description and not limitation. Hence, although the present disclosure has been described hereinabove by way of illustrative embodiments thereof, it may be modified, without departing from the spirit, scope and nature of the subject disclosure.
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/CA2012/000736 | 8/6/2012 | WO | 00 | 4/10/2014 |
| Number | Date | Country | |
|---|---|---|---|
| 61521022 | Aug 2011 | US |
