EP2977113A1

EP2977113A1 - CMUT ultrasound focusing by means of partially removed curved substrate

Info

Publication number: EP2977113A1
Application number: EP14178245.8A
Authority: EP
Inventors: designation of the inventor has not yet been filed The
Original assignee: Koninklijke Philips NV
Current assignee: Koninklijke Philips NV
Priority date: 2014-07-24
Filing date: 2014-07-24
Publication date: 2016-01-27

Abstract

A curved array of micromachined transducers (MUTs, 17) is arranged on an outer (6) surface of a substrate (15). In accordance to the invention, a curvature profile of the curved array is defined by a residual stress distribution within the substrate (15). The substrate (15) further comprises at least two layers, wherein the residual stress distribution within the substrate (15) may be determined by at least two layers of which a first layer (21) has a first thermal expansion coefficient value (e1); and a second layer (22) has a second thermal expansion coefficient value (e2), which differs from the first thermal expansion coefficient value.

Description

FIELD OF THE INVENTION

This invention relates to a curved array of ultrasound capacitive micromachined transducers (CMUTs) arranged on an outer surface of a substrate. Further this invention relates to an ultrasound probe comprising a curved array of ultrasound transducers. Further this invention relates to a manufacturing method of providing such curved array.

BACKGROUND OF THE INVENTION

For 1D sensor arrays, focusing can be done by electronic circuitry means in the azimuth direction, while the elevation focus is accomplished through spatially positioning sensors along elevation direction with a small axial offset, namely by curving the flat array. Another way of achieving the elevation focusing is applying a focusing property in the acoustic lens. This is often implemented in 2D arrays having a fixed focal point.
A curved sensor device, such as an ultrasonic transducer array, is fabricated from a flat micromachined sensor (such as capacitive micromachined transducer: CMUT or piezoelectric micromachined transducer: pMUT) array constructed using micromachined electromechanical system (MEMS) techniques is disclosed in US20050146247 A1 . This device comprises: a backing layer having a rear surface that adopts a curved profile and a multiplicity of sensors built on the front surface of the backing layer. The rear surface of the backing layer can be bent forward or backward and attached to a curved front face of a support member, thereby causing the sensors to adopt a curved array.
The drawback of the known curved array is complexity in the manufacturing steps requiring separate steps of manufacturing of the multiplicity of sensors on the flat substrate; bonding this substrate to the backing layer; partial dicing through the substrate and the backing layer and attaching the backing layer to the curved support member.
It is therefore desirable to provide an improved curved array of sensor devices, in particular ultrasound transducers.

SUMMARY OF THE INVENTION

It is an object of present invention to provide a curved array of ultrasound transducers (which can be manufactured simpler), wherein the plurality of the MUT transducers is fully integrated into the curved substrate.
According to the invention this object is realized by defining a curvature profile of the curved array through a residual stress distribution within the substrate.
Usually the residual stress in thin layers of different materials constituting MEMS causes strain and is an undesirable effect in device preparation. In contrast, the present invention exploits this property and suggests applying the residual stress distribution in the substrate in order to provide a defined curvature profile to the substrate. Since a plurality of MUTs (CMUTs or pMUTs) in array can be manufactured on a substrate's surface using known manufacturing methods, the controlled residual stress in the substrate can provide a controlled curvature profile to the array for a deliberate focusing or defocusing of the ultrasound waves.
In an embodiment of the present invention the substrate comprises at least two layers wherein the residual stress distribution within the substrate is determined by at least two layers of which a first layer has a first thermal expansion coefficient value; and a second layer, which is coupled to the first layer and located on the outer surface side, has a second thermal expansion coefficient value, which differs from the first thermal expansion coefficient value.
This embodiment describes a particular realization of the residual stress distribution via providing the substrate comprising two layers of materials having different thermal expansion coefficient values. The differences in expansion coefficients between materials of the substrate layers results in different expansion of these layers at room temperatures. A predefined selection of layers having different thermal expansion coefficient values permits determining the desired residual stress in the substrate at the end of an ultrasound array fabrication.
In an embodiment of the present invention the first thermal expansion coefficient value is bigger than the second thermal expansion coefficient value resulting at room temperature in a convex shape of the curvature profile of the substrate at the outer surface side.
Due to the fact that the first layer has bigger thermal expansion coefficient value compared to the second layer, the first layer constituting the inner surface of the substrate will tend to shrink more than the second layer at room temperature. This results into a compressive residual stress in the substrate at the outer surface side. The atoms of the material of the second layer would be experiencing a negative force, which pulls them closer than they would be in a bulk state. The consequence of this homogenous compressive stress is a curving of the substrate outwards, wherein the curvature profile takes a convex shape at the outer surface side. Since the plurality of the CMUTs is arranged on the same surface side, the transmitted or received acoustic waves would have a negative axial focus.
In another embodiment of the present invention the first thermal expansion coefficient value is smaller than the second thermal expansion coefficient value resulting at room temperature in a concave shape of the curvature profile of the substrate at the outer surface side.
Here, due to the fact that the first layer has smaller thermal expansion coefficient value compared to the second layer, the first layer constituting the inner surface of the substrate will tend to shrink less at the given temperature than the second layer located at the outer surface side. This results into a tensile residual stress in the substrate at the outer surface side. The atoms of the material of the second layer would be experiencing a positive force, which pulls them farther apart than they would be in a bulk state. The consequence of a homogenous tensile stress is a curving of the substrate into a concave shape at the outer surface side. Since the plurality of the CMUTs is arranged on the same surface side, the transmitted or received acoustic waves would have a positive axial focus.
Yet in another embodiment of the present invention the first layer silicon oxide having a first thickness and the second layer is silicon having a second thickness, wherein the first thickness is smaller than the second thickness.
The introduction of the thickness' difference between the layers gives an additional control over a curvature radius of the concave shaped curved array. Thinning the second layer may give lower curvature radius, thus bringing the focus distance of the acoustic waves closer to the array's surface.
In yet another embodiment of the present invention the curvature profile of the curved array comprises at least a concave area of the concave shape and a convex area of the convex shape at the outer surface side; wherein the concave area is located in the center of the curved array along at least one of the lateral and elevation directions and has a smaller curvature radius than the radius of the convex area, which is located at the edges of the curved array along at least one of the lateral directions.
The residual stress distribution can be adapted to cause a change of a sign of the curvature radius throughout the curvature profile. In other words, the substrate's outer surface can vary its curvature from the concave to the convex shapes. This embodiment covers the case when the central part of the curved array has the concave shape and the edges of the array take a convex shape. This may be implemented in both one-dimensional (1D) and two-dimensional (2D) arrays, wherein the change in the curvature happens either along one direction: the only direction in 1D array and one of the two directions in 2D array; or along both lateral directions of the 2D array. The advantage of this embodiment is improving of the curved array focusing characteristics; wherein the concave area has the positive axial focus and the edge area has the negative focus permitting the reception and transmission of acoustic signals beyond the focal point of the concave area.
In yet another embodiment the concave area is located at the edges of the curved array along at least one of the lateral directions and has a bigger curvature radius than the radius of the convex area, which is located in the center of the curved array along at least one of the lateral and elevation directions.
This embodiment shows further control of the curved array focusing characteristics; wherein the center convex area has the negative axial focus and the edge area has the positive focus permitting the reception and transmission of acoustic signals beyond the focal point of the concave area at the edges.
In a further embodiment of the present invention first layer comprises:

a middle region (Smin) of a minimum thickness (dmin) located in the center of the curved array along the lateral direction and having a middle region length (lm); and
a gradient region (Sgr) having a gradual thickness change from a maximum thickness (dmax) at the edge of the array down to the minimum thickness (dmin) in the middle region.

This embodiment explores an opportunity of providing the change of the curvature profile shape from the concave in the center of the array to the convex towards the edges through an introduction of different thickness regions in the first layer, wherein the thickness value varies along the lateral direction of the array. Reducing the first layer thickness in the middle region causes the change of sign of the residual stress within this region compared to the array edges. The abruptness of the residual sign changes is determined by the surface area (or length in 1D array) of the middle region and the surface area (or length in 1D array) of the gradient region. Thus, controlling the thicknesses of these regions gives a control over the convex and concave areas curvature; this specific embodiment allows creating a so called "mexican hat" curvature profile.
In yet further embodiment of the invention the curved array further comprises an integrated circuitry electrically coupled to the transducers and adapted to transmit and receive focused ultrasound beams.
The integrated circuitry is electrically coupled to the curved array of the present invention and provides a possibility for focusing transmitted and received ultrasound signals.
These and other aspects of the invention will be apparent from and elucidated with reference to the embodiments described hereinafter.

BRIEF DESCRIPTION OF THE DRAWINGS

In the drawings:

FIGURE 1 illustrates a curved CMUT array arranged on an outer surface of a substrate;
FIGURE 2 illustrates a CMUT cell of the CMUT array,
FIGURE 3 depicts a coordinate system of the array according to the present invention;
FIGURE 4 (a) illustrates a double layer substrate of the curved array in case of the zero residual stress, wherein the first coefficient value e1 of the first layer is bigger than the coefficient e2 of the second one; (b) illustrates the developed tensile residual stress in the substrate (c) illustrated a concave shape curvature profile of the curved array resulted from the tensile residual stress at room temperature; (d) illustrates a double layer substrate of the curved array in case of the zero residual stress, wherein the first coefficient value e1 of the first layer is smaller than the coefficient e2 of the second one; (e) illustrates the developed compressive residual stress in the substrate (f) illustrated a convex shape curvature profile of the curved array resulted from the compressive residual stress at room temperature;
FIGURE 5 illustrates an estimation of the curvature radius from the curvature profile;
FIGURE 6 depicts simulations of the curvature profile of the curved array of the curved array along positive values of the x-axis according to one of the embodiments of the present invention;
FIGURE 7 depicts simulations of the curvature profile of the curved array along positive values of the x-axis according another embodiment of the present invention;
FIGURE 8 illustrates a tapering process of the substrate occurring during RIE;
FIGURE 9 (a) illustrates different areas of the substrate having a "Mexican hat" curvature profile; (b) and (c) illustrate particular embodiments of the thickness variations of the middle region in order to control the "Mexican hat" curvature profile;
FIGURE 10 depicts simulations of the "Mexican hat" curvature profile of the curved array along positive values of the x-axis according to yet another embodiment of the present invention;
FIGURE 11 depicts simulations of the "Mexican hat" curvature profile of the curved array along positive values of the x-axis according to further embodiment of the present invention; and
FIGURE 12 illustrates in block diagram form an ultrasonic imaging system arranged to be operated in accordance with the principles of the present invention.

DETAILED DESCRIPTION OF EMBODIMENTS

FIGURE 1 shows an embodiment of a curved 1D ultrasound array comprising a substrate 15 having an outer 6 and inner 5 surfaces and a plurality of capacitive micromachined transducers (CMUTs) 17 arranged on the outer 6 surface of the substrate 15.
An illustrative representation of a CMUT cell constituting the CMUT 10' array is presented in FIGURE 2. A membrane or diaphragm 114 is suspended above a floor 130 of the CMUT cell with a gap 8 there between. A membrane electrode 7 is coupled to the cell membrane 114 and can move with the membrane 114. In this embodiment a substrate electrode 7' is embedded into the floor 130 of the cell located at the outer surface 6 of the substrate 15. Other realizations of the electrode 7 design can be considered, such as electrode 7 may be embedded in the membrane 114 or it may be deposited on the membrane 114 as an additional layer. In this example, the substrate electrode 7' is circularly configured and embedded into the cell floor 130. In addition, the membrane layer 114 is fixed relative to the cell floor 130 and configured and dimensioned so as to define a spherical or cylindrical cavity 8 between the membrane layer 114 and the cell floor 130. The cell floor 130 may comprises CMOS compatible materials.
The cell and its cavity 8 may have alternative geometries. For example, cavity 8 could define a rectangular or square cross-section, a hexagonal cross-section, an elliptical cross-section, or an irregular cross-section.
The substrate electrode 7' is typically insulated on its cavity-facing surface with an additional insulating layer (not shown). A material for the insulating layer can be silicon oxide-nitride-oxide (ONO), high-k dielectrics and oxides (various grades including silane, SiH4, based PECVD SiO2). The insulating layer may advantageously reduce charge accumulation which leads to device instability and drift and reduction in acoustic output pressure. Use of the insulating layer is desirable with CMUTs with collapsed membrane. This type of CMUT is more susceptible to charge retention than CMUTs operated with suspended membranes. The disclosed components may be fabricated from CMOS compatible materials, e.g., Al, Ti, nitrides (e.g., silicon nitride), oxides (various grades: thermal or TEOS/SiH4 LPCVD/PECVD based), poly-silicon and the like. In a CMOS fabrication process, for example, the oxide and nitride layers may be formed by chemical vapor deposition and the metallization (electrode) layer put down by a sputtering process. Suitable CMOS processes are LPCVD and PECVD, the latter having a relatively low operating temperature of less than 400°C. Exemplary techniques for producing the disclosed cavity 8 involve defining the cavity in an initial portion of the membrane layer 114 before adding a top face of the membrane layer 114. In the exemplary embodiment depicted in FIGURE 2, the diameter of the cavity 8 may be larger than the diameter of the second electrode 7'. The membrane electrode 7 may have the same outer diameter as the substrate electrode 7, although such conformance is not required.
The CMUT fabrication process can comprise either the "sacrificial release process", wherein the cavity underneath of the membrane is formed by first applying a sacrificial layer on the substrate, then applying the membrane layer followed by the removing of the sacrificial layer with a selective etchant; or the "wafer bonding process", wherein the cavity is formed on the prime wafer and the membrane in another wafer, then both wafers are bonded together such as the cavity confined by the membrane is formed (B.T. Khuri-Yakub, J. Micromech. Microeng. 21 (2011) 054004).
The membrane 7 and substrate 7' electrodes of the CMUT cell provide the capacitive plates of the CMUT device and a gap of the cavity 8 form the dielectric between the plates of the capacitor.
The membrane electrode 7 can be brought in vibration by means of a signal transmitter/receiver 105 adapted to apply an AC and DC voltage over the substrate and membrane electrodes, which result in the generation of an acoustic beam. When later on the membrane vibrates as a result of the received acoustic signal, the changing dimension of the dielectric gap between the electrodes leads to changing capacitance of the CMUT which is detected by the signal transmitter/receiver 105 as the response of the CMUT cell to a received acoustic echo. The CMUT cell may comprise additional electrodes integrated either in the membrane 114 or/and the substrate 15 (or the cell floor 130) for separate or additional AC/DC voltage supply provided for the cell driving. Electrical connection to the CMUT device, often by means of an integrated circuit (IC) such as an application specific integrated circuit (ASIC) facilitates both transmission and reception modes of the device. ASIC may also comprise integrated the signal transmitter/receiver 105 module. The cell can also comprise a layer(s) 32 of an acoustic window or lens material comprising the acoustic coupling, protective or focusing properties.
ASIC electrically coupled to the curved CMUT array can also facilitate a partial (micro) or full beamforming function, which permits steering and focusing transmitted and received ultrasound signals.
In accordance with the principles of the present invention a stress distribution is introduced within the substrate 15, such that the ultrasound transducer array 10' is curved and its curvature profile is defined by a residual stress distribution. In further embodiments, we will describe a one-dimensional curved ultrasound array, wherein the described curvature profile is defined by the array's cross section in a lateral plane (xz), as shown in FIGURE 3. However, it shall be understood by the person skilled in the art that the principles of the present intention can be also applied to a two-dimensional array.
Thermal mismatch stress is the more common source of a residual stress. Stress or strain commonly exist in thin films (layers) as a result of constraints imposed by their substrates. Stress (σ[Pascal]) is the force per unit area that is acting on a surface of solid. Strain (ε) is a measure of the deformation of a solid proportional to stress being experienced by this solid. A thin film and its substrate generally have different thermal expansion coefficients, so stress is produced during temperature changes occurring in deposition and annealing. A piece of solid is under stress when its atoms are displaced from their equilibrium positions by a force. The displacement is governed by the interatomic potential (Micromachined Thin-Film Sensors for SOI-CMOS Co-Integration Ch.2:Thin dielectric films stress extraction, J. Laconte, D. Frandre, J.-P Raskin, Springer, 2006, ISBN 978-0-387-28842-0). An external tensile force tends to lengthen the solid and in turn to increase the interatomic distance. A force which increases the interatomic distance is positive, and hence the tensile force (or stress) is positive. An external compressive force (or stress) which tends to ·shorten the interatomic distance in solid is negative.
In a first embodiment of the present application a residual stress distribution can be created by intrinsic stress difference in the internal structure of the material. The development of a stress gradient from the inner 5 to the outer 6 substrate's surfaces permits inducing either tensile or compressive residual stress at the outer surface side. For example, polysilicon doping with phosphorous atoms creates a region of more compressive stress than polysilicon. In the meantime, boron doping of silicon introduces a tensile stress when introduced into the crystal lattice. As the smaller boron atom displaces the silicon atom, there is a tendency for the lattice to contract locally, therefore resulting in local tensile stress. Depending on the residual stress gradient in the substrate, it can bend, i.e. adapt a curvature profile. An induced change in the crystallinity of the same material can also introduce residual stress and may be applicable to practice the present invention. In the second embodiment of the present invention, the residual stress distribution in the substrate 15 can be controlled by providing a multilayer substrate. During the manufacturing process the layers are deposited onto each other at elevated temperatures. Assuming thermal equilibrium between the layers, the deposition process at the elevated temperatures is stress free. The difference in thermal expansion coefficient will result in a residual stress once the substrate is brought to room temperatures; depending on the residual stress distribution the substrate will bend, i.e. adapt a curvature profile.
In FIGURE 4 the example of providing the curved array arranged on the outer surface of a double layer substrate is shown. The double layer substrate 15 comprises a first layer 21 constituting the inner surface of the substrate and having a first thermal coefficient value e1; and a second layer 22 constituting the outer surface of the substrate and having a second thermal coefficient value e2.
The Stoney equation allows determining a relation between the residual stress value in the second layer and a curvature radius R of the profile: $σ_{2} = \frac{d_{1}^{2} E_{1}}{6 (1 - v_{1})} \frac{1}{d_{2}} \frac{1}{R}, [N / m^{2}]$

wherein E₁ is Young's modulus of the first layer, v₁ is the Poisson ratio of the first layer, d₁ and d₂ are thicknesses of the first and the second layers correspondingly. For a bi-axial substrate, the stress σ is related to the strain ε (deformation) via the biaxial modulus E/(1-v) as σ=E/(1-v)*ε.
The residual stress in the second layer as result of a mismatch in thermal expansion coefficient can be also described by: $σ 2 = \frac{E}{1 - v} (e 2 - e 1) Δ T,$

wherein ΔT is the difference in between the elevated temperature of the layer deposition process and room temperature, which has a positive value in this definition. Similar equation can be written for the first layer. Thus, the thermal mismatch in the expansion coefficients leads to the stress value in the given layer.
FIGURES 4A to 4C illustrate a case of the tensile residual stress. This can be achieved by selecting materials of the first 21 and the second 22 layers such that the first thermal coefficient value e1 is smaller than the second one e2. At the end of the manufacturing process the first layer constituting the inner surface 5 of the substrate will tend to shrink less (directions of the acting forces are shown with arrows) than the second layer at the given temperature (FIGURE 4B). This results into a tensile residual stress in the substrate at the outer surface 6 side. The atoms of the material of the second layer would be experiencing a positive force, which pulls them farther apart than they would be in a bulk state. The consequence of a homogenous tensile stress is a curving of the substrate into a concave shape of the curvature profile at the outer surface 6 side. Since the plurality of the CMUTs is arranged on the same surface side, the transmitted or received acoustic waves would have a positive axial focus, e.i. transmitted or received ultrasound waves have a focal point on the outer surface 6 side of the substrate 15 corresponding to the positive values of z-axis (FIGURE 3).
FIGURES 4D to 4F illustrate a case of the compressive residual stress. The compressive residual stress can be achieved by selecting materials of the first 21 and the second 22 layers such that the first thermal coefficient value e1 is bigger than the second one e2. At the end of the manufacturing process the first layer will tend to shrink more than the second layer at the given temperature of the fabrication step. This results into a compressive residual stress in the substrate at the outer surface side. The atoms of the material of the second layer would be experiencing a negative force (FIGURE 4E), which pulls them closer than they would be in a bulk state. The consequence of this homogenous compressive stress is a curving of the substrate outwards (FIGURE 4F), wherein the curvature profile takes a convex shape at the outer surface side.
It shall be understood that the CMUT ultrasound array 7 arranged on the substrate's outer surface may also have its own residual stress (which may be mainly dominated by the membrane layer stress). The value of this residual stress would depend on the exact fabrication process and the CMUT's cell design.
Controlling the exact ratio between coefficients of thermal expansion (e1 and e2) provides a control over the curvature radius of the ultrasound curved array in the lateral (xz)-plane. The radius, R, can be estimated as shown in FIGURE 5 as R=(L/2)²/(2h_max), wherein L - is a length of 1D array (along elevation direction in FIGURE 3) and h_max - is a maximal deflection of the substrate's surface. It is chosen throughout the application that for the concave shape the value of R is positive, while for the convex shape R is negative. Note, for illustration purpose the directions of the inner 5 and outer surfaces of the substrate are also shown in FIGURE 3. The same orientation of the curved array is used for the simulations described below.
The introduction of the thickness' difference (Stoney equation) between the layers of the substrate 15 gives an additional control over a curvature radius of the curved array. The thinner the layer of the substrate material compared to the rest of the substrate the stronger its expansion will be influenced by the presence of other thicker layers and their expansion coefficients. This can be understood from the point of view, that layer (film) thickness defines the amount of atomic layers away from interlayer interface. A bigger amount of atomic layers within the material layer provides greater material volume for the stress relaxation developed at the interface due to a mismatch of atom positions in different materials.
The Stoney equation is also applicable in the case of the constant layer thickness. In case, the thickness of the layers changes throughout the substrate a finite model simulations shall be performed in order to predict the actual curvature profile. As can be seen from the Stoney equation, in case of fixed parameters of the materials the curvature radius becomes smaller as the first layer thickness is reduced or the second layer thickness is increased. The exact shape of the curvature would depend on the thermal coefficient's ratio between the two layers.
FIGURE 6 illustrates a simulation of the curvature profile for positive x-coordinate values in the lateral (xz)-plane distribution in the concave shaped curved array. The center of the x-axis coincides with the center of the curved array in the lateral direction. The curvature profile of the array is symmetric around the z-axis. The lateral size of the simulated curved array is 25 cm, while the lateral dimension of the array in the figure is normalized, the substrate comprises the first layer of silicon material, which is 400 micrometer thick and the second layer of thermal silicon oxide, which is 2 micrometer thick. This example, illustrates a relatively small residual stress towards the outer surface is tensile, causing a relatively (micrometer-range compared to the array's dimension) small deflection (displacement) of the substrates outer surface , which results in the concave shaped curvature profile with a positive curvature radius of around 5 m.
FIGURE 7 illustrates a simulation of similar case when the thickness of the first layer of the array in previous embodiment is reduced to 50 micrometer and the second layer thickness remains the same. This substrate configuration of the curved array provides negative positive curvature radius of about 11 cm, almost an order of magnitude lower than in previous example . For the simulations, following mechanical properties were used: silicon (100) - Young's modulus 130 GPa and Poisson ratio 0.28; thermal silicon oxide - Young's modulus 70 GPa and Poisson ratio 0.20.
Thinning the second layer may give smaller curvature radius, thus bringing the focus distance of the acoustic waves closer to the array's surface.
The non-homogenous residual stress distribution gives access to vary the sign of the curvature radius throughout the curvature profile.
In the next embodiment, the non-homogenous residual stress distribution is realized via changing the thickness (introducing a thickness gradient) of the first layer 21 along the lateral direction. The thickness gradient in the silicon layer, for example, can be achieved with a standard (deep) reactive ion etching technique ((D)RIE), by making use of a process of tapering in the silicon. FIGURE 8 illustrates the tapering process occurring during the RIE of a silicon layer 40. During the etching the polymer resist layer 41, commonly used to protect the designated area 42 from ion etching, is being etched during the etching process as well, thereby reducing (direction of the etching is indicated with arrows) the designated area 42 size over time as shown in FIGURE 8B and 8C. Usually, the etching rate of the resist layer is lower than silicon etching rate and this phenomenon is normally exploited to tune the angle of the edge 43 of the layer (flatten with the etching time) in order to get a good step coverage for the next layer. The exact parameters of the reactive ion etching process, such as gas pressure (typically maintained in a range between a few microbar and a few hundred microbar) and discharge voltage used to create ions, determine the exact rates of the materials etching. Another way to control the angle of the edge 43 is varying the thickness of the resist layer 41.
FIGURE 9A shows one of the realizations of this embodiment in accordance with the present invention. The curved array 17 has the curvature profile with a convex area 51 of the convex shape and two concave areas 52 of the convex shape at the outer surface side 6. The convex area 51 has a negative curvature radius of R_convex, while the concave area has a negative curvature radius R_concave. The convex area 51 is located in the center of the curved array along the lateral direction and has a smaller curvature radius than the radii of the concave areas, which are located at the edges of the curved array along the lateral directions. Technologically this type of curved area can be realized via reducing the thickness of the first layer 21 of the substrate 15. As has been explained above the process of tapering (thinning down) of the first material layer can be achieved through (D)RIE.
The following manufacturing steps can be applied:

providing the substrate comprising the outer 6 and inner 5 surfaces; wherein the inner surface is introduced by the first layer having a first thermal coefficient value e1 and the maximum thickness (d_max); and the outer surface is introduced by the second layer 22 constituting the substrate and having a second thermal coefficient value e2.
providing an array of the CMUT cells arranged on the outer surface 6 side of the substrate;
providing a resist layer atop the inner surface of the substrate; wherein the resist layer outlines a middle (or center) region (Smin), leaving the middle region exposed to an etchant;
thinning the thickness of the first layer 21 of the substrate 15 in the middle region Smin down to a minimum thickness (dmin) as illustrated in (FIGURE 9B). The thinning step via (D)RIE would also introduce an gradient thickness region Sgr with a gradual thickness change in the ranges in between of dmax and dmin.

After the removal of the resists layer at room temperature the residual stress distribution would cause the curved array to curve the array's profile creating the concave and the convex areas. Defining the etching process parameters, the resists thickness and size and position of the outlined middle region gives control over the radii of the concave and convex areas. The residual stress distribution can be adapted to cause a change of a sign of the curvature radius throughout the curvature profile. In other words, the substrate outer surface can vary its curvature from the concave to the convex shapes.
In a particular embodiment shown in FIGURE 9C the middle region Smin is located in the center of the curved array along the lateral direction; the gradient region Sgr is located in between the middle region and the array's edge. The residual stress distribution leads to the central part of the curved array having the convex shape and the edges of the array having the concave shape. When the R_convex of the region in the center is smaller than R_concave of the region at the edges of the array, the curvature profile is called a "Mexican hat". This may be implemented in both one-dimensional (1D) and two-dimensional (2D) arrays, wherein the change in the curvature happens either along one direction: either perpendicular to the lateral direction in 1D array (elevation focus) or along the lateral direction (axial focus); and one of the two directions in 2D array; or along both lateral directions of the 1D and 2D arrays. The advantage of this embodiment is improving of the curved array focusing characteristics; wherein the concave area has the positive axial focus and the edge area has the negative focus permitting the reception and transmission of acoustic signals beyond the focal point of the concave area.
FIGURE 10 illustrates a simulation of curvature profile in the "Mexican hat" type of curved array illustrated in FIGURE 9C. The lateral size of the curved array is again 25 cm, the substrate comprises the first layer of silicon material. The parameters of the silicon layer are: the middle region is located in the middle of the array and has a middle region length of 12.5 cm (two times smaller than the lateral size of the array) and the minimum thickness of 50 micrometer; while at the edges the fisrt layer has the maximum thickness of 100 micrometer (two times bigger than dmin). The second layer of the substrate is thermal silicon oxide with thickness of 2 micrometer. The residual stress changes its sign from the center towards the edges and causes the middle region to adapt a convex shape with a negative curvature radius of around 21 cm.
FIGURE 11 illustrates a simulation of similar case which differs from the previous case by the maximum thickness value, which is increased to 400 micrometer, and the length of the of the middle region and the minimum thickness value, which are zero (the thickness of the first layer changes from zero in the middle up to 400 micrometer at the edges). This substrate configuration of the curved array provides the negative curvature radius of the convex area of around 15 cm.
For most (3D) applications a 1D symmetric (around the curvature axis) array is preferred. Though it shall be understood by the skilled in art person that this invention can be applied to other arrays ('Biplanar' for example), wherein the different radii of curvature along the lateral and elevation axis can be used, especially if both planes have different properties, for example focal distance or different operating frequencies. Such curved 2D array would allow variation in resolution/focal distance or combining treatment with imaging functionalities (low freq HIFU and high frequency imaging).
Referring to FIGURE 12, an ultrasonic diagnostic imaging system with a CMUT array probe 10 is shown in block diagram form. The curved ultrasound transducer array 10' is either 1D or 2D array of CMUT elements capable of scanning in a 2D plane or in three dimensions for 3D imaging. In case of 3D imaging and sometimes in 2D imaging the transducer array is coupled to a microbeamformer 12 in the probe which controls transmission and reception of signals by the CMUT array cells. Microbeamformers are capable of at least partial beamforming of the signals received by groups or "patches" of transducer elements as described in US Pats. 5,997,479 (Savord et al.), 6,013,032 (Savord ), and 6,623,432 (Powers et al.) The microbeamformer is coupled by the probe cable to a transmit/receive (T/R) switch 16 which switches between transmission and reception modes. The transmission of ultrasonic beams from the transducer array 10 under control of the microbeamformer 12 is directed by a transducer controller 18 coupled to the T/R switch and the main system beamformer 20, which receives input from the user's operation of the user interface or control panel 38. One of the functions controlled by the transducer controller is the direction in which beams are steered and focused. The transducer controller 18 can be coupled to control a DC bias control 45 for the CMUT array 10'. The DC bias control 45 controls the signal transmitter/receiver(s) 105 of the CMUT array can be coupled to the beamformer 20 directly or through the microbeamformer 12.
During reception the partially beamformed signals produced by the microbeamformer 12 (in case it is used) are coupled to a main beamformer 20 where partially beamformed signals from individual patches of transducer elements are combined into a fully beamformed signal.
In case of 1D array the microbeamformer 12 and beamformer 20 are replaced by the mechanical-former, which realized the mechanical steering of the array in the elevation direction.
The beamformed signals are coupled to a signal processor 22. The signal processor 22 can process the received echo signals in various ways, such as bandpass filtering, decimation, I and Q component separation, and harmonic signal separation which acts to separate linear and nonlinear signals so as to enable the identification of nonlinear (higher harmonics of the fundamental frequency) echo signals returned from tissue and microbubbles. The signal processor may also perform additional signal enhancement such as speckle reduction, signal compounding, and noise elimination.
The processed signals are a scan converter 32 and a multiplanar reformatter 44. The scan converter arranges the echo signals in the spatial relationship from which they were received in a desired image format. For instance, the scan converter may arrange the echo signal into a two dimensional (2D) sector-shaped format, or a pyramidal three dimensional (3D) image. The scan converter can overlay a B mode structural image with colors corresponding to motion at points in the image field corresponding with their Doppler-estimated velocities to produce a color Doppler image which depicts the motion of tissue and blood flow in the image field. The multiplanar reformatter will convert echoes which are received from points in a common plane in a volumetric region of the body into an ultrasonic image of that plane, as described in US Pat. 6,443,896 (Detmer ). A volume renderer 42 converts the echo signals of a 3D data set into a projected 3D image as viewed from a given reference point as described in US Pat. 6,530,885 (Entrekin et al. ) The 2D or 3D images are coupled from the scan converter 32, multiplanar reformatter 44, and volume renderer 42 to an image processor 30 for further enhancement, buffering and temporary storage for display on an image display 40. A graphics processor 36 generates graphic overlays for display with the ultrasound images. These graphic overlays can contain standard identifying information such as patient name, date and time of the image, imaging parameters, and the like. For these purposes the graphics processor receives input from the user interface 38, such as a typed patient name. The user interface is also coupled to the transmit controller 18 to control the generation of ultrasound signals from the transducer array 10' and hence the images produced by the transducer array and the ultrasound system. The user interface is also coupled to the multiplanar reformatter 44 for selection and control of the planes of multiple multiplanar reformatted (MPR) images which may be used to perform quantified measures in the image field of the MPR images.
While the invention has been illustrated and described in detail in the drawings and foregoing description, such illustration and description are to be considered illustrative or exemplary and not restrictive; the invention is not limited to the disclosed embodiments. Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims.
In the claims, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality. A single element or other unit may fulfill the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.

Claims

A curved array of micromachined transducers (MUTs, 17) arranged on an outer (6) surface of a substrate (15),
characterized in that
a curvature profile of the curved array is defined by a residual stress distribution within the substrate (15).
The curved array according to claim 1, further comprising an integrated circuitry coupled to the transducers and adapted to transmit and/or receive ultrasound beams.
The curved array according to claim 1 and 2, wherein the substrate (15) comprises at least two layers wherein the residual stress distribution within the substrate (15) is determined by at least two layers of which a first layer (21) has a first thermal expansion coefficient value (e1); and a second layer (22) has a second thermal expansion coefficient value (e2), which differs from the first thermal expansion coefficient value.
The curved array according to claim 3, wherein the first thermal expansion coefficient value is bigger than the second thermal expansion coefficient value resulting at room temperature in a convex shape of the curvature profile of the substrate at the outer surface side.
The curved array according to claim 3, wherein the first layer is silicon oxide having a first thickness (d1) and the second layer is silicon having a second thickness (d2), wherein the first thickness is smaller than the second thickness.
The curved array according to claim 5, wherein the first thickness (d1) is at least 10 times smaller than the second thickness (d2).
The curved array according to claim 5, wherein the first thickness (d1) is at least 50 or 200 times smaller than the second thickness (d2).
The curved array according to claim 2, wherein the first thermal expansion coefficient value is smaller than the second thermal expansion coefficient value resulting at room temperature in a concave shape of the curvature profile of the substrate at the outer surface side.
The curved array according to claim 3, wherein the curvature profile of the curved array comprises at least a concave area (52) of the concave shape and a convex area (51) of the convex shape at the outer surface side (6).
The curved array according to claim 9, wherein the concave area is located at the edges of the curved array along at least one of lateral directions and has a bigger curvature radius than the radius of the convex area, which is located in the center of the curved array along at least one of the lateral and elevation directions.
The curved array according to claim 10, wherein the first layer comprises:
- a middle region (Smin) of a minimum thickness (dmin) located in the center of the curved array along the lateral direction and having a middle region length (lm); and

- a gradient region (Sgr) having a gradual thickness change from a maximum thickness (dmax) at the edge of the array down to the minimum thickness (dmin) in the middle region.
The curved array according to claim 10, wherein the middle area length is at least two times smaller than a lateral size of the array.
The curved array of ultrasound transducers according to claim 12, wherein the middle region length is zero and the minimum thickness of the middle region is zero.
An ultrasound probe comprising:
- the curved array of the ultrasound transducers according to any of claim 1 to 13;

- electronic circuitry coupled to the curved array to activate the transducers to generate ultrasound beams;

- a beamforming means coupled to the circuitry and adapted to steer the ultrasound beams.
A manufacturing method of a curved array of micromachined transducers (MUTs, 17) comprising:
- providing a substrate (15);

- arranging the array of MUTs on an outer (6) surface of the substrate,
characterized in that the method further comprises
defining a residual stress distribution within the substrate (15), wherein the definition of the residual stress provides a curvature profile of the curved array.