WO2012014010A1

WO2012014010A1 - System and method for operating capacitive micromachined ultrasonic transducers

Info

Publication number: WO2012014010A1
Application number: PCT/IB2010/053379
Authority: WO
Inventors: Selim Olcum; Yalcin Yamaner; Ayhan Bozkurt; Hayrettin Koymen; Abdullah Atalar
Original assignee: Selim Olcum; Yalcin Yamaner; Ayhan Bozkurt; Hayrettin Koymen; Abdullah Atalar
Priority date: 2010-07-26
Filing date: 2010-07-26
Publication date: 2012-02-02

Abstract

The present invention relates to capacitive micromachined ultrasonic transducers (CMUTs) and describes a novel method and system for their operation in collapse and release mode of the membrane. The system comprising; a CMUT cell which is the active acoustical part of a transducer array or an element, a membrane which is the moving part of the CMUT cell under acoustical or electrical excitation, a substrate which is the surface on which the fabrication processes are being conducted and the final CMUT transducers are being working, a top electrode which is the moving electrode embedded to the moving membrane of the CMUT cell, a bottom electrode which is the second electrode of the CMUT cell and an insulation layer which is an insulating material between the top electrode and the bottom electrode, a gap which is the vacuum cavity between the substrate and the membrane and a voltage source generating the drive voltage signal between the top and the bottom electrode.

Description

SYSTEM AND METHOD FOR OPERATING CAPACITIVE MICROMACHINED ULTRASONIC TRANSDUCERS

BRIEF DESCRIPTION OF THE INVENTION

5

The present invention relates to capacitive micromachined ultrasonic transducers (CMUTs) and describes a novel method and system for their operation in collapse and release mode of the membrane.

10 BACKGROUND OF THE INVENTION

Capacitive micromachined ultrasonic transducers have a suspended active membrane for acoustic transduction. The suspended membrane may be fabricated using a wafer bonding process as disclosed in the U.S. Patent No. US2004/0085858

15 or can be released using a sacrificial layer process as disclosed in the U.S. Patent No. US2005/0177045. The active membrane is excited using an electrode which is embedded to the membrane. The electrode can be fabricated on the top of the membrane, buried in the membrane or at the bottom of the membrane. Top electrodes with different coverage of the membrane were reported in the previous

20 publications or patents. The top electrode can be compromised of plurality of electrodes to be excited separately for providing high acoustic pressures without collapse of the membrane as disclosed in the U.S. Patent No. US2005/0200241.

The U.S. Patent No. US2009/0322181 has disclosed the active top electrode can be positioned at the sides of the membrane leaving the center empty. This way the effect of charging in the insulation material is minimized while achieving high pressure output. The disadvantage of this embodiment, however, is it utilizes less electrical force as compared to a CMUT with complete electrode coverage on the membrane and substrate for the same operating voltage.

30 The second electrode of the CMUT is a stationary electrode which is located on the substrate. A metal layer can be fabricated on the substrate as the stationary electrode or the substrate itself can be used as the electrode if the wafer has high conductivity. The International Patent No. WO2009/016606 has disclosed the bottom electrode which is compromised with two separate electrodes. This way with the electrode at the center, the membrane is collapsed and the electrodes at the sides are used to oscillate the membrane for transmission.

The electrical insulation of the moving and stationary electrodes is achieved by the use of an insulation layer such as silicon oxide, silicon nitride or other insulation materials. The insulation layer can be used on top of the bottom electrode, or under the top electrode or both.

The sizes of the CMUT membranes typically vary from a few microns to a few millimeters in radius and from a few hundred nanometers to a few hundred microns in thickness. Typical CMUT transducers comprise the plurality of the CMUT cells connected in parallel or in phased array configuration. Utilized electrical signals vary between a few volts to a few hundred volts. CMUTs are fabricated in several different ways utilizing surface micromachined or wafer bonding technology, which were disclosed extensively in the previous reports of publications or patents.

The moving membrane of a CMUT is activated by a DC voltage applied between the two electrodes. If an AC voltage is applied on either of the electrodes, the membrane starts to vibrate and emits acoustic signals. If at the activated state, an acoustic pressure signal excites the membrane, an AC voltage is generated between the electrodes due to the capacitance change. A similar operation is also true in the case of an electrical or pressure pulse excitation. The conventional operation is achieved when the membrane does not touch the substrate in either transmit or receive cycles as disclosed in the U.S. Patent No. US5894452. At this operation regime, when the electrical force generated between the moving and stationary electrodes at a specified bias voltage, is larger than the restoring force of the membrane, the membrane collapses to the substrate. The smallest voltage that is enough for the membrane to collapse is called the collapse voltage of the membrane. Once the membrane collapses, reducing the voltage would not result an imminent release of the membrane. The smallest voltage that would keep the membrane in contact with the substrate is called the snap-back voltage. Typically snap-back voltage is less than the collapse voltage, which would result in the hysteresis behavior of the CMUT membranes that was demonstrated in the previous studies and publications.

The U.S. Patent Nos. US7274623 and US2005/0219953 describe different regimes of operation where the membrane is brought in contact with the substrate. In the operation regime which is disclosed in the U.S. Patent No. US2005/0219953, the pulse amplitude combined with the applied DC voltage is always higher than the snap-back voltage. It is said that at this operation region the center of the membrane remains in contact with the substrate. However, this may not be true during the pulse cycle, if the membrane is loaded with a high enough potential energy, the membrane can snap-back even if the DC potential on the membrane is higher than the snap- back voltage.

The U.S. Patent No. US7274623 has disclosed the operation regime which has two different operation of collapse mode. The membrane is kept in contact with the substrate initially, then released and pulled back in the first mode. In the second mode, the released membrane is brought in contact with the substrate and released back to the initial position.

It is true that the membrane incorporates a higher electromechanical coupling efficiency once the membrane is in contact with the substrate; however, if the purpose is to maximize the pressure output of the CMUTs, the definition of the collapse mode is not adequate. There is a need of operating CMUTs to generate the maximum available acoustic pressures at specified frequency spectrum using the specified maximum available voltage. The transducer dimensions and the electrical signals must be designed accordingly in order to meet the acoustic pressure/spectrum demands of the ultrasonic industry.

Other operation modes are introduced incorporating the membrane to touch the substrate disclosed in Patent Nos. US2006/0004289 and WO2009/073562. These modes use two separate modes of the membrane movements. In the first mode the membrane is used as a free vibrating spring, and in the second mode the membrane is collapsed on to a dimple structure and utilizes a smaller gap region for higher sensitivity.

The methods and structures disclosed in the referenced prior art, lack the description of how high pressure outputs can be achieved using capacitive micromachined ultrasonic transducers with a bandwidth around a specific frequency. Therefore there is a need in the art to specify the following parameters for high pressure pulse transmission;

the coverage area of the bottom or top electrode (single or multiple each) over the membrane,

the thickness of the insulation layer between the top and the bottom electrodes for a given maximum voltage,

- the gap height for a given center frequency of operation,

radius and thickness of the membrane,

the value of the collapse voltage for a fixed amount of applied voltage, rise and fall times of the applied electrical signal,

the width of the applied electrical pulse signal.

OBJECTS AND SUMMARY OF THE INVENTION

It is a general object of the invention to design and operate CMUTs such that high acoustic signal is generated at the output of the transducer when they are excited by a rising or falling voltage step signal. It is a further object of the invention to design the output signal's rise and fall times, for achieving a high acoustic pressure output when CMUTs are excited by a rising or falling voltage step signal. It is a further object of the invention to maximize the output acoustic pressure by tuning the delay time between the rising and falling voltage steps applied to CMUTs.

It is a further object of the invention to generate acoustic pulses with desired durations while maintaining high acoustic pressures, using the rising or falling voltage steps.

It is a further object of the invention to generate maximum peak-to-peak pressure signal, when a rising and falling voltage step follow each other with a delay at either order. The pressure output is maximized by optimizing the delay between the falling and rising voltage steps. The amount of the delay may be different for different devices, different voltages or different order of voltage steps.

It is a further object of the invention to generate maximum peak-to-peak pressure signal by optimizing the rise and fall times of the applied voltage steps. The voltage fall time during the releasing edge should be made as small as possible, independent of the device configuration or voltage. The output pressure amplitude of the membrane during the rising voltage step is less prone to the rise time of the voltage step; however it is better to make it small.

The invention utilizes voltages applied to the membrane more than the collapse voltage of the membrane. In the case of a fixed maximum available voltage either because of the driving circuit limitations or fabrication process limitations, the thinnest insulation layer is chosen between the active electrodes, which will withstand the maximum applied electric field. The invention includes the usage of electrodes with coverage of more than 80%. The output pressure increases as the electrode coverage increases.

The invention is utilized to a method where a membrane is collapsed or released using a voltage step. The acoustical output of the membrane at each step is maximized by the invention. The gap height is chosen such that the desired pulse duration is achieved at the output. A collapse voltage is chosen for the CMUT such that a maximum pressure output is achieved when the CMUTs are driven with the designed electrical signal. The radius and the thickness of the membrane are determined with respect to the collapse voltage need and the fabrication limitations.

DETAILED DESCRIPTION OF THE INVENTION

System and method for operating of capacitive micromachined ultrasonic transducers realized in order to fulfill the objects of the present invention is illustrated in the attached figures, where:

Figure 1 shows the cross section of a collapsed CMUT cell in the preferred embodiment of the invention.

Figure 2 shows the restoring and electrostatic forces of a circular CMUT cell with respect to average displacement of the membrane.

Figure 3 shows the delivered energy to the medium when the CMUT cell is released and collapsed with respect to the applied voltage step. The inset shows a zoomed view of the collapse point.

Figure 4 schematically shows the FEM model of a circular CMUT cell used in the FEM simulations.

Figure 5 shows the comparison of the emitted average pressure by a CMUT cell calculated by FEM simulation to experimental result of the same. Figure 6 schematically shows the equivalent circuit model used in the SPICE simulations. Figure 7 shows the comparison of the emitted average pressure by a CMUT cell calculated by SPICE simulation to experimental and FEM results of the same.

Figure 8 shows the FEM simulation results of the peak-to-peak pressures when the CMUT cells are excited by a release/collapse cycle, with respect to the amplitude of the voltage step for full electrode coverage (squares) and half electrode coverage (diamonds).

Figure 9 shows the equivalent circuit simulation results of the negative pressure amplitude of the transmitted acoustic signal when the CMUT cells are excited by a collapsing voltage step, with respect to the amplitude of the step for different insulation layer thicknesses.

Figure 10 shows the equivalent circuit simulation results of the positive pressure amplitude of the transmitted acoustic signal when the CMUT cells are excited by a releasing voltage step, with respect to the amplitude of the step for different insulation layer thicknesses.

Figure 11 shows the equivalent circuit simulation results of the full width at half maximum of the transmitted negative acoustic signal when the CMUT cells are excited by a collapsing voltage step, with respect to the amplitude of the step for different insulation layer thicknesses.

Figure 12 shows the equivalent circuit simulation results of the full width at half maximum of the transmitted positive acoustic signal when the CMUT cells are excited by a releasing voltage step, with respect to the amplitude of the step for different insulation layer thicknesses. Figure 13 shows the peak pressure of the transmitted acoustic signal when the CMUT cells are excited by a collapsing (dashed) and releasing (solid) voltage steps, with respect to the rise (dashed) and fall (solid) times of the applied step. Figure 14 shows the peak-to-peak pressure amplitude of the transmitted acoustic signal when the CMUT cells are excited by a collapse/release voltage steps with respect to the amount of the delay between the applied step signals for voltage amplitudes of 100V (dashed) and 200V (solid). Figures 15 and 16 show the peak-to-peak pressure amplitude of the transmitted acoustic signal when the CMUT cells are excited by a collapse/release voltage steps of 200V in amplitude, with the optimum delay time. The pressure levels are depicted as a function of the membrane dimensions in terms of a/t_m, and for different values of collapse voltages. In the simulations the gap height is chosen to be A) 200nm and B) 1 OOnm.

Figures 17 and 18 show the peak-to-peak pressure amplitude of the transmitted acoustic signal when the CMUT cells are excited by a collapse/release voltage steps of 100V in amplitude, with the optimum delay time. The pressure levels are depicted as a function of the membrane dimensions in terms of a/t_m, and for different values of collapse voltages. In the simulations the gap height is chosen to be E) 200nm and G) lOOnm.

Figures 19 and 20 show the center frequency of the transmitted acoustic signal when the CMUT cells are excited by a collapse/release voltage steps of 200V in amplitude, with the optimum delay time. The pressure levels are depicted as a function of the membrane dimensions in terms of a/t_m, and for different values of collapse voltages In the simulations the gap height is chosen to be B) 200nm and D) lOOnm.

Figures 21 and 22 show the center frequency of the transmitted acoustic signal when the CMUT cells are excited by a collapse/release voltage steps of 200V in amplitude, with the optimum delay time. The pressure levels are depicted as a function of the membrane dimensions in terms of a/t_m, and for different values of collapse voltages In the simulations the gap height is chosen to be B) 200nm and D) lOOnm.

Figures 23 and 24 show the peak-to-peak pressure amplitude of the transmitted acoustic signal when the CMUT cells are excited by a collapse/release voltage steps of A) 200V B) 100V, in amplitude, with the optimum delay time. The pressure levels are depicted as a function of the membrane dimensions in terms of a/t_m, and for different values of gap heights.

Figures 25 and 26 show the center frequency of the transmitted acoustic signal when the CMUT cells are excited by a collapse/release voltage steps of A) 200V B) 100V, in amplitude, with the optimum delay time. The pressure levels are depicted as a function of the membrane dimensions in terms of a/t_m, and for different values of gap heights.

Figures 27 and 28 show the peak-to-peak pressure amplitude of the transmitted acoustic signal when the CMUT cells are excited by a collapse/release voltage steps with different amplitudes and with the optimum delay time. The pressure levels are depicted as a function of the membrane dimensions in terms of a/t_mj and for different values of collapse voltages in the simulations the gap height is chosen to be A) lOOnm and C) 200nm. Insulation thickness is adjusted with respect to the maximum voltage amplitude.

Figures 29 and 30 show the center frequency of the transmitted acoustic signal when the CMUT cells are excited by a collapse/release voltage steps with different amplitudes and with the optimum delay time. The pressure levels are depicted as a function of the membrane dimensions in terms of a/t_m, and for different values of collapse voltages in the simulations the gap height is chosen to be A) lOOnm and C) 200nm. Insulation thickness is adjusted with respect to the maximum voltage amplitude. Figure 31 shows the peak-to-peak pressure amplitude of the transmitted acoustic signal when the CMUT cells are excited by a collapse/release voltage steps with different amplitudes and with the optimum delay time. The pressure levels are depicted as a function of the membrane dimensions in terms of a/t_m, and for different values of gap heights and voltage amplitudes. Insulation thickness is adjusted with respect to the maximum voltage amplitude. The collapse voltages of the simulated devices are adjusted such that the voltage step amplitude is always 10 times of the collapse voltage.

Figure 32 shows the center frequency of the transmitted acoustic signal when the CMUT cells are excited by a collapse/release voltage steps with different amplitudes and with the optimum delay time. The pressure levels are depicted as a function of the membrane dimensions in terms of a/t_mj and for different values of gap heights and voltage amplitudes. Insulation thickness is adjusted with respect to the maximum voltage amplitude. The collapse voltages of the simulated devices are adjusted such that the voltage step amplitude is always 10 times of the collapse voltage. Figure 33 shows the average pressure as a function of time emitted from a CMUT cell for a collapse/release voltage steps in opposite orders. The amplitude of the voltage steps is 200V and the delay between the two steps is chosen to be optimum for maximum pressure transmission. Figure 34 shows the pressure waveform transmitted by a CMUT designed for operation around 10 MHz using 100V. The dimensions of the designed CMUT are 81 μιη radius, 5 μιη thickness, 100 nm gap and 200 nm insulator thickness. The top electrodes cover the full surface of the active region. The applied electrical pulse has a 100V amplitude, 10 ns rise and fall times and 22 ns width.

Figure 35 shows the frequency spectrum of the pressure waveform depicted in FIG. 34. CMUT designed for operation around 10 MHz using 100V. Figure 36 shows the pressure waveform transmitted by a CMUT designed for operation around 25 MHz using 100V. The dimensions of the designed CMUT are 42 μιη radius, 2.3 μιη thickness, 70 nm gap and 200 nm insulator thickness. The top electrodes cover the full surface of the active region. The applied electrical pulse has a 100V amplitude, 10 ns rise and fall times and 10 ns width.

Figure 37 shows the frequency spectrum of the pressure waveform depicted in FIG. 36. CMUT designed for operation around 25 MHz using 100V. Figure 38 shows the pressure waveform transmitted by a CMUT designed for operation around 5 MHz using 200V. The dimensions of the designed CMUT are 158 μιη radius, 9.3 μιη thickness, 200 nm gap and 400 nm insulator thickness. The top electrodes cover the full surface of the active region. The applied electrical pulse has a 200V amplitude, 10 ns rise and fall times and 48 ns width.

Figure 39 shows the frequency spectrum of the pressure waveform depicted in FIG. 38. CMUT designed for operation around 5 MHz using 200V.

Elements shown in the figures are numbered as follows:

1. System for operating of capacitive micromachined ultrasonic transducers

2. CMUT cell

3. Membrane

4. Substrate

5. Gap

6. Top electrode

7. Bottom electrode

8. Insulation layer

9. Voltage Source

System (1) for operating capacitive micromachined ultrasonic transducers (CMUT) is composed of; at least one CMUT cell (2) which is the active acoustical part of a CMUT transducer array or an element or a single cell transducer. A CMUT cell (2) may consist of single or several elements. Each element consists of single or several membranes (3).

at least one suspended membrane (3) which is the moving part of the CMUT cell (2) under acoustical or electrical excitation. The membrane (3) corresponding to a collapse voltage to collapse to the substrate (4) and to a snapback voltage to resume its suspended position over the gap (5).

at least one substrate (4) which is the surface on which the fabrication processes are being conducted and the final CMUT transducers are being working. The substrate (4) may be a highly or poorly conductive silicon wafer or any other type glass or insulating wafers suitable for microfabrication processes.

at least one gap (5) between the substrate (4) and the membrane (3).

at least one top electrode (6) which is the moving electrode (6) embedded to the moving membrane (3) of the CMUT cell (2) and has a more than 80% coverage of the membrane (3) including the center of the moving membrane (3). Top electrode (6) may be fabricated on top of the membrane (3), buried in the membrane (3) or at the bottom of the membrane (3). The moving top electrode (6) may be fabricated as the sole electrode (6) on the membrane (3) or may be divided into two or more parts.

at least one bottom electrode (7) which is the second electrode (7) of the CMUT cell (2) and which is the stationary electrode (7) embedded to a substrate (4), and has more than 80% coverage including center of the cell (2) under the moving region. A highly conductive metal layer can be used as the bottom electrode (7) or else highly conductive substrate (4) can be used as the bottom electrode (7).

at least one insulation layer (8) which is an insulating material between the top electrode (6) and the bottom electrode (7) and has a breakdown voltage of Vbrk- The insulation layer (8) material can be silicon nitride, silicon oxide or similar dielectric materials deposited or grown between the top electrode (6) and the bottom electrode (7). The insulation layer (8) can be a part of the moving membrane (3) or else can be deposited on top of the stationary bottom electrode (7) or else both.

the membrane (3) corresponding to a collapse voltage to collapse to the substrate (4) and to a snapback voltage to resume its suspended position over the gap (5),

the gap (5), which is vacuum cavity between the substrate (4) and the membrane (3).

at least one voltage source (9) for applying a drive voltage signal between the electrodes (6 and 7), with an amplitude greater than 5 times the collapse voltage.

In the preferred embodiment of the invention, the voltage source (9) is capable of generating a drive voltage signal for applying a bias voltage between the electrodes (6 and 7) with an amplitude between zero and the collapse voltage, keeping the membrane (3) in suspended position, and thereafter the voltage source (9) applies a drive voltage between the electrodes (6 and 7), giving a rising voltage step that would bring the voltage level to V_max,, which is greater than 5 times the collapse voltage, to cause the membrane (3) to collapse, and after a time delay, the voltage source (9) gives a falling voltage step between the electrodes (6 and 7) that would bring the final voltage to a value between zero and the snap-back voltage to cause the membrane (3) to return its suspended position over the gap (5). The parameter 5 times the collapse voltage is a result obtained by the analysis and simulations performed on the CMUT cell (2) and will be discussed in the following parts of this document.

In an alternative embodiment, the voltage source (9) is capable of generating a drive voltage signal for applying a bias voltage between the electrodes (6 and 7) with an amplitude, V_max, which is greater than 5 times the collapse voltage, keeping membrane (3) in the collapsed position, and thereafter, the voltage source (9) applies a drive voltage between the electrodes (6 and 7), giving a falling voltage step_, which brings the voltage level between zero and the snap-back voltage to cause the membrane (3) to snap-back, and after a time delay said voltage source (9) gives a rising voltage step between the electrodes (6 and 7) that would bring the final voltage to V_max to cause the membrane (3) to return its initial collapsed position. The parameter 5 times the collapse voltage is a result obtained by the analysis and simulations performed on the CMUT cell (2) and will be discussed in the following parts of this document

A cross section of a collapsed, circular capacitive micromachined ultrasonic transducer (CMUT) cell (2) is depicted in FIG. 1. In a preferred embodiment of the invention, the CMUT cell (2) has circular shape; however the shape of the CMUT cell (2) (rectangular, hexagonal, etc.) does not affect the results.

The radius and thickness of the CMUT cell's (2) membrane (3) are defined as a and t_m respectively. The thickness of the insulation layer (8) between the top electrode (6) and the bottom electrode (7) is defined as t_t. The membrane (3) is suspended on the substrate (4) with a height of t_g.

In a preferred embodiment of the invention, the membrane (3) is made of silicon nitride. However the material of the membrane (3) is not critical for the analysis and the results, therefore in alternative embodiments the membrane (3) can be made of silicon.

In a preferred embodiment of the invention the substrate (4) is made of high conductive silicon material however any other type of glass or insulating wafers suitable for microfabrication processes can be used and do not affect the results.

The insulation layer (8) is a part of the membrane (3) in a preferred embodiment, however in the case of an alternative embodiment; the insulation layer (8) can be deposited on top of the bottom electrode (7), which would not affect the performance and the results of this disclosure. Yet in an alternative embodiment the insulation layer (8) can be used both at the bottom of the membrane (3) and on top of the bottom electrode (7). The radiated pressure output from a CMUT cell (2) when it is excited by an electrical signal is related to the forces acting on the membrane (3). The CMUT cell

(2) can be excited such that it undergoes two distinctive mechanical movements: collapse and release. The static membrane (3) defiection profile at any DC bias point is required for determining the electrical and mechanical forces acting on the membrane (3). The restoring force of a membrane (3) can be calculated by simply multiplying the uniform pressure deflecting the membrane (3) with the area. Any applied pressure is balanced by the restoring force of the membrane (3). The electrical force applied on the membrane (3) under any voltage excitation can be calculated by half the derivative of the capacitance with respect to average displacement times the applied voltage squared. In these calculations; any defiection profile caused by the applied voltage can be approximated by the deflection caused by a uniform pressure was assumed. This is a good assumption when the membrane

(3) has both the top electrode (6) and the bottom electrode (7) overlapping over the full surface of the membrane (3), which will be clarified in the following parts, is the optimal coverage for maximum power transmission.

The change in the electrical and restoring forces with respect to the average displacement when a CMUT cell (2) is excited by a 200V bias is depicted in FIG. 2. The intersection point of the electrical force with mechanical restoring force determines the equilibrium displacement point. The displacement is calculated as 132.7nm, which is a very good approximation to the FEM simulation result of 133.7 nm for the device dimensions in FIG. 2. The RMS error between the deflection profiles of the analytical result and the electrostatic FEM simulation is 0.6%.

The mechanical energy stored by the membrane (3) is released to the immersion medium when the voltage across the membrane (3) is pulled back to zero. Stored energy can be calculated by integrating the restoring force curve to find the area of the delivered energy (collapsed) region of FIG. 2 and is found as 0.63 nJ for the CMUT cell (2) of interest. The average displacement and the mechanical force have a linear relation at the conventional region, where the center of the membrane (3) does not touch the substrate (4). When the membrane (3) touches the substrate (4), the relation becomes highly nonlinear and the amount of energy stored in the membrane (3) increases faster compared to the conventional region of operation. When a high voltage is applied across the membrane (3), a part of the input electrical energy is stored as the mechanical energy while another part is delivered to the immersion medium. Since the net force applied to the medium is the difference between the electrical and restoring force curves, the energy transferred to the medium can be found by the area between the two curves. This area corresponding to 2 nJ of energy is shown as the delivered energy (released) region of in FIG. 2.

Energies delivered to the liquid medium in which the CMUT cell (2) radiates during the collapse and the release periods are shown in FIG. 3 in two separate curves as a function of applied voltage. Clearly, more energy is delivered during the collapse period as compared to release period. The difference increases as the applied voltage is increased further. This asymmetry does not exist for the conventional uncollapsed operation when the membrane (3) acts like a linear spring.

Increasing the voltage further than the collapse voltage increases the radiated energy in both collapse and release periods drastically as seen in FIG. 3. Here, it should be noted that these results are true if the top electrode (6) is covering the full surface of the membrane (3). For the case of a partial top electrode's (6) or bottom electrode's (6) coverage, the membrane (3) cannot store more energy once the contact radius is equal to the electrode (6 or 7) radius. Increasing the delivered energy would improve the power delivered to the medium, but there is no simple relation between them. In order to determine the power, the dynamic problem involving the membrane (3) mass, the radiation impedance of the immersion medium and the nonlinear spring constant of the membrane (3) excited by a deflection dependent force is need to be solved. This problem has been solved by performing finite element method simulations and SPICE model simulations. The finite element simulations were done using ANSYS Multiphysics Environment. A 2D axisymmetric model of CMUT cell (2) is created as shown in FIG. 4. To simulate the CMUT cell (2) operation, a coupled electrostatic-structural analysis was performed. The electrostatic environment was modeled using PLANE 121 elements. The structural environment was created using PLANE82 elements and using CONTACT 172-TARGET 169 pair elements that can simulate the contact in the collapse operation. The membrane (3) is modeled with 2-D 8-Node Structural Solid (PLANE82) elements. Electromechanical elements (TRANS 126) were generated under the bottom surface nodes of the membrane (3) using the ANSYS built-in macro EMTGEN'. The macro requires a gap (5) value "GAP" to generate ground plane nodes under the selected nodes and creates TRANS 126 elements in between. It also performs a point-wise capacitance calculation and provides the necessary inputs for each TRANS 126 element. The GAPMIN" parameter defines the maximum possible deflection before contact. A contact stiffness factor, FKN=\ is used to overcome convergence problems with a reasonable penetration at contact interface. Material parameters of the membrane (3) used in the FEM simulations are given in TABLE 1. Effect of the atmospheric pressure is included in all simulations by applying a constant pressure of 0.1 MPa on the top surface nodes of the membrane (3). Dynamic behavior of a CMUT cell (2) is simulated using the same model with a fluid loading. A fluid column is created over the membrane (3) using 2-D axisymmetric harmonic acoustic fluid (FLUID29) elements. Coupling of the structural motion to the fluid pressure at the interface is enabled by specifying fluid- structure flags. The fluid column height is set to a large value ensuring that there is no reflection from the top boundary at the end of the simulation. The height of the fluid column is set to 2 mm. Transient effects are turned off in the first step to ensure a stable membrane (3) under the DC bias. Afterwards, transient effects are turned on and the analysis is performed for 1 ISQC . The average pressure is captured at 1 mm above the membrane (3) surface. The accuracy of the FEM simulations are tested by comparing the FEM results to experiments. In FIG. 5, the measured acoustic pulse emitted from the membrane (3) is depicted when it is excited by a minus 140V pulse and 140 V bias, along with the FEM simulation result. The FEM simulation results are larger in amplitude and faster in time due to the errors in the radiation impedance. The rigid baffle defined for the simulations, enforces a high radiation resistance and zero imaginary part. However in the case of a real transducer array, the radiation impedance seen by the array has a lower real part and a nonzero and positive imaginary part.

TABLE 1: Material parameters used in the FEM and SPICE simulations

The effect of the electrode (6 and 7) coverage on the output pressure is tested using the FEM simulations and experiments. In FIG. 8, the results of the FEM simulations show that maximum pressure from a CMUT cell (2) can be achieved using full top and bottom electrode (6 and 7) coverage. The same result is achieved from the experimental results. However, electrode (6 and 7) coverages can be smaller than 100% of the active membrane (3) region and the output pressures can still be high enough if the electrodes (6 and 7) cover the center region and the remaining uncovered parts are at the edges, where there is little contribution to electrostatic forces. Here it should be put a limit to these uncovered regions, which is 20% according to FEM simulations. Therefore the electrodes (6 and 7) should cover at least 80% of the active membrane (3) region including the center region where electrostatic forces are maximum. In the rest of this document, all the CMUT cells (2) are assumed to have a full coverage of top and bottom electrode (6 and 7) with respect to the membrane (3) area.

For an alternative and faster way to FEM simulations, the dynamic problem of a CMUT cell (2) under large signal excitation is solved using SPICE model seen in FIG 6. The model included the force curves of a CMUT cell (2) with respect to the average displacement. These forces, as a function of average displacement, are used as voltage sources (9) in the equivalent circuit model. The membrane (3) as a spring is inherently represented in the restoring force and the membrane (3) as a mass is modeled by an inductor with a value of 1.8 times the membrane (3) mass. The radiation impedance is assumed to be real and pc for a sufficiently large array, where p is the density and c is the velocity of sound in the immersion medium.

Static average deflections calculated by the circuit are within 1% of the ANSYS simulation results. It demonstrated that the use of the equivalent circuit model by comparing its pulse response results to FEM simulations and immersion experiments. As seen in the FIG. 7, the SPICE simulation results of the equivalent circuit is in good agreement with the results of FEM and the experiment.

One critical parameter for the CMUT cell (2) is the insulation layer (8) between the electrodes (6 and 7), which should be thick enough for withstanding the high electric fields, and should be thin enough for maintaining most of the field in the vacuum gap (5). In this analysis the insulation layer (8) is chosen to be silicon nitride, which has a dielectric breakdown field of -1000 V/μιη. For the sake of safety, during the following analysis, the thickness of the silicon nitride insulation layer (8) is chosen such that the electric field in the layer does not exceed 500 V/μιη. The following analysis and results does not change if other material or maximum operating fields are chosen. However the maximum operating fields should not be less than half of the dielectric strength for high acoustic transmission, and obviously it cannot be more than the dielectric strength of the insulation layer (8). In other words, if the voltage source (9) generating a drive voltage signal has a maximum voltage, V_max, then the dielectric breakdown voltage, Vbrk of the insulation layer (8) should be more than V_max and less than two times V_max. In FIG. 9 and FIG. 10, negative and positive pressure pulses calculated by SPICE simulations as a function of rising and falling voltage steps are depicted, respectively. Different curves in the figures represent different insulation layers (8). For a thicker insulation layer (8), higher maximum voltage amplitude is used. Increasing the voltage amplitude for achieving higher pressure outputs is also demonstrated in FIG. 3. However if the amount of voltage to be applied to the CMUT cells (2) is being limited, the insulation layer (8) thickness must be chosen accordingly. For example, if there is a voltage limit of 200V, the insulation layer (8) thickness, t; must be at most 0.4μιη for maximum pressure output. The value of the insulation layer (8) thickness should be chosen during the design stage and according to the maximum available voltage and maximum allowable electric field. For this example, 0.4 μιη thickness can be easily calculated using the maximum voltage of 200V divided by the half the dielectric strength value which is 500 V/μιη. In the following sections, the t; values for the simulated CMUT cells (2) are calculated using this rule.

Before analyzing the dynamics of the CMUT cells (2), the effect of the rise and fall times of the voltage steps have been determined by using SPICE simulations. The results of the simulations are depicted in FIG. 13. +200V and -200V steps are applied to CMUT cells (2) with varying rise and fall times. The dimensions of the CMUT cell (2) under consideration has 30 μτα radius, 1 μτα thickness, 0.2 μτα gap (5) and 0.4 μτα insulation layer (8) with full electrode (6 and 7) coverage on the membrane (3). A +200V rising step would cause the membrane (3) to collapse and it is called as the collapsing edge or rising voltage step. The rise time of the collapsing edge is not very critical up to 40 ns in this example. A reasonable rise time for the collapsing edge can be chosen for high pressure output. A -200V falling step would cause the membrane (3) to be released and it is called as the releasing edge or falling voltage step. The fall time of the releasing edge is very prominent and should be chosen very short regardless of the resonance frequency of the membrane (3). The results are depicted in FIG 13. In the remaining part of this description, the rise and fall times of the pulses are chosen as 10ns, for the sake of simplicity. It should be noted here that faster fall times would result higher pressures. Therefore the voltage source (9) that would generate the driving voltage signal as a falling voltage step should have fall time such that the output pressure of the CMUT cell (2) is maximum. Similarly the voltage source (9) that would generate the driving voltage signal as a rising voltage step should have rise time such that the output pressure of the CMUT cell (2) is maximum

The rising and falling voltage steps defined in the previous paragraph are usually used together when generating an acoustic pulse. The rising and falling voltage steps applied to CMUT cells (2) may be in either order. However both cases require an optimal delay for maximum pressure generation. In FIG. 14, extracted peak-to-peak pressure values from two different CMUT cells (2) using FEM simulations are depicted with respect to the amount of delay between the rising and falling voltage steps. In all the simulation results reported in remaining part of this description, the optimum amount of delay is calculated and used by the SPICE equivalent circuit. Therefore the voltage source (9) that would generate the driving voltage signal with consecutive rising and falling steps in either order should have a time delay between them such that output pressure of the CMUT cell (2) is maximum.

The order of the rising and falling voltage steps is critical for the pulse shape. In FIG. 33, two different pressure pulse shapes are depicted, which are obtained by driving the CMUT cell (2) with the same pulse amplitude, but in opposite collapse/release order. For the release step, the stored mechanical energy is radiated into the medium as a positive pressure waveform. The membrane (3) is accelerated by high restoring forces and the force acting on the membrane (3) decreases as the membrane (3) is released. Once the membrane (3) transfers its energy to the medium its velocity drops. Therefore a damped waveform is transmitted into the medium. However, in the collapse voltage step, the electrical forces are low at first and increases with the displacement. Therefore, at the DC stable point the membrane (3) still has high velocity and kinetic energy, which result in an underdamped waveform and ringing. Therefore a collapse/release cycle would result in a better behaving waveform as seen in FIG. 33. In addition the excess energy at the end of the collapse step can be used advantageously for the release step, if the delay time is chosen optimally. This extra energy would result in higher acoustic pressure levels for the collapse/release cycle. At the rest of the description, collapse/release cycle is used during the simulations; however, a release/collapse cycle would result in similar pressure levels and center frequencies. For the collapse-release cycle, which means the membrane (3) is initially unexcited. First it is collapsed using a rising voltage step and released back to initial position by a falling voltage step after a time delay. During this operation, the voltage value of the initial voltage is taken as zero for the simulations and experiments. However this value can be larger than zero and its effect on the pressure output would be minimal as long as it is less than the collapse voltage. Obviously, the smaller this initial voltage is the more pressure the CMUT cell (2) would generate. Through-out this document this initial value is taken as zero, but it can be smaller than the collapse voltage of the membrane (3) and still generate large pressures. Therefore the voltage source (9) that would generate the collapse/release cycle gives a drive voltage signal by applying a bias voltage between the electrodes (6 and 7) with an amplitude between zero and the collapse voltage, keeping the membrane (3) in suspended position, and thereafter the voltage source (9) applies a drive voltage between the electrodes (6 and 7), giving a rising voltage step that would bring the voltage level to V_max,, which is greater than 5 times the collapse voltage, to cause the membrane (3) to collapse, and after a time delay, the voltage source (9) giving a falling voltage step between the electrodes (6 and 7) that would bring the final voltage to a value between zero and the snap-back to cause the membrane (3) to return its suspended position. The parameter 5 times the collapse voltage is a result obtained by the analysis and simulations performed on the CMUT cell (2) and will be discussed in the following parts of this document Similarly, during a release/collapse cycle first, the membrane (3) is released using a falling voltage step and collapsed back to initial position by a rising voltage step after a time delay. During this operation, the voltage value that would keep the membrane (3) in released position is taken to be zero for the simulations and experiments. However this value can be larger than zero and its effect on the pressure output would be minimal as long as it is less than the snap-back voltage. Obviously, the smaller this voltage is the more pressure the CMUT cell (2) would generate. Therefore, the voltage source (9) that would generate the release/collapse cycle gives a drive voltage signal by applying a bias voltage between the electrodes (6 and 7) with an amplitude, V_max, which is greater than 5 times the collapse voltage, keeping the membrane (3) in the collapsed position, and thereafter, the voltage source (9) applies a drive voltage between the electrodes (6 and 7), giving a falling voltage step, which brings the voltage level between zero and the snap-back voltage to cause the membrane (3) to snap-back, and after a time delay the voltage source (9) gives a rising voltage step between the electrodes (6 and 7) that would bring the final voltage to V_max to cause the membrane (3) to return its initial collapsed position. The parameter 5 times the collapse voltage is a result obtained by the analysis and simulations performed on the CMUT cell (2) and will be discussed in the following parts of this document

The collapse mode of operation has been demonstrated at the previous studies and patents in terms of collapse and collapse-snap back operations. However, the dependency of the output pressure on the level/depth of collapse has not been explored. In this invention, a parameter γ, which will represent the depth of collapse, is defined, γ is simply the ratio of the amplitude of applied voltage (collapsing or releasing) to the collapse voltage of the membrane (3). For the voltage source (9) that would generate the driving voltage signals on the electrodes (6 and 7) of the CMUT cell (2), the amplitude of the applied voltage is previously defined as V_max. Therefore the γ parameter is simply V_max divided by the collapse voltage. A higher γ would imply a higher contact radius during the collapse stage. In FIG. 15, the effect of the parameter γ on the peak-to-peak output pressure of different CMUT cells (2) is depicted. The applied signal to the CMUT cells (2) in all the simulations is a 200V rising voltage step followed by a -200V falling voltage step with an optimum amount of delay. The rise and fall times of the steps are chosen to be constant at 10ns, which is also true for the rest of this description. The insulation layer's (8) thickness, is chosen to be 400nm, because of the maximum voltage amplitude 200V and half of the dielectric strength which is 500 V/μιη. For different curves in FIG. 15, different values of γ are used. As seen in FIG. 15 the maximum peak-to- peak pressure that can be achieved is higher for higher γ values. However the peak- to-peak pressure values are also dependent on the membrane (3) dimensions. FIG. 15 would imply that, for achieving higher output pressures, one should use higher γ values. However output pressure amplitudes do not change much after the γ value of 5. For values less than 5, output pressure of a CMUT cell (2) drops quickly. Since γ is the ratio of V_maxto collapse voltage, V_max, which is drive voltage signal amplitude of the voltage source (9) should be greater than 5 times the collapse voltage. One other result of FIG. 15 is the dimensions of the membrane (3) should be chosen such that maximum pressure output can be achieved.

In most cases, high pressure output is not enough to define an application specific requirement. The center frequency of the generated pulse is also an important parameter to be considered. In FIG. 19, the center frequency of the generated pulses of FIG. 15 is depicted as a function of the membrane (3) dimensions and for different γ. One obvious result is the center frequency of the pressure pulse increases with the increasing pressure output. However the center frequency changes faster than the pressure output, which is useful for designing CMUT cells (2) generating a high pressure pulse at a desired center frequency. The same plots are made in FIGS. 16, 17, 18, 19, 20 and 22 for pulse amplitudes of 200V, 100V and 100V, and gap (5) heights of lOOnm, 200nm and lOOnm, respectively. In the FIGS. 15 to 22 it is seen that the effect of the gap (5) is very prominent on the adjustment of center frequency of the generated pulse. The change of the peak-to- peak pressure and center frequency with respect to the membrane (3) dimensions for different gap (5) heights is depicted in FIG. 23 and FIG. 25 respectively. The maximum voltage amplitude is assumed to be 200V and the insulation layer (8) thickness is chosen accordingly. As the gap (5) height increases, the amplitude and the center frequency of the pulse decreases. The same analysis is performed when the maximum voltage amplitude is 100V and results are depicted in FIG. 24 and FIG. 26. A similar trade off between the pressure amplitude and the center frequency can be achieved by tuning the maximum available voltage. In FIG. 27 and FIG 29, the peak-to-peak pressure amplitude and center frequency is depicted with respect to the membrane (3) dimensions for different maximum available voltages. The gap (5) height is assumed to be constant and lOOnm. The same plots are made for 200nm gap (5) in FIG. 28 and FIG. 30. Pressure pulses with lower center frequencies can be achieved by decreasing the voltage, however the pressure amplitude drops more than when the center frequency is tuned by the gap (5) height. It is always better use higher voltages for high pressure outputs and higher gaps (5) for lower frequency pulses.

One may require very high amplitude pressure pulses with low center frequencies. This time one should increase the maximum available voltage and the gap (5) height at the same time. In FIG. 31 and FIG. 32, pressure amplitude and center frequency of the generated pulses for different gap (5) height and voltages. This way it is possible to maintain high acoustic pressures for lower frequencies at the expense of high voltages. Those designs would require thick insulation layers (8) or materials with higher dielectric breakdown.

The method for achieving a high pressure output from CMUT cells (2) with a specified frequency band can be best described by the use of examples using the results presented in previous paragraphs. This way a method is carried out for designing the dimension and electrical signals to be applied for specific ultrasound applications.

As a first example, it is supposed that there is a need of a pressure pulse signal with as high as possible amplitude with a spectrum around 10 MHz and using maximum 100V. The transients of applied electrical signal should be as fast as possible for higher output pressures as depicted in FIG. 13 is determined. For this specific example it is managed to achieve rise and fall times of 10 ns for 100V rising and falling voltage steps. Using the information provided in FIG. 33, it is chosen to use a collapse-release cycle for this case. A full electrode (6 and 7) coverage for both top and bottom electrodes (6 and 7) which would generate the highest possible pressure amplitude for a given voltage as depicted in FIG. 8 is chosen. The maximum voltage to be applied is limited by the requirements and is 100V. So, the insulation layer's (8) thickness is determined by keeping in mind the results of FIG. 9 and 10. The thickness is chosen to be 200 nm, which is believed it would have a dielectric breakdown of more than 100V.

After determining the insulation layer's (8) thickness, the gap (5) height of the membranes (3) for achieving the high pressure output at the desired frequency spectrum is should be chosen. Using FIGS. 21 and 22 it is seen that 200 nm of gap (5) is too large for achieving 10 MHz center frequency. Choosing smaller gaps (5) if possible increases the electrical forces acting on the membrane (3) and in return increases the pressure output. Therefore for this specific example the gap (5) height is chosen to be 100 nm.

Using FIGS. 18 and 22 the a/t_m ratio is chosen 16. Using FIGS. 18 and 22, it is found out that; in order to achieve a center frequency of 10 MHz and maintain high pressure amplitude; the gamma parameter (maximum voltage divided by the collapse voltage) should be 13. After choosing the gamma parameter value to be 13 and the radius-thickness ratio to be 16 the collapse voltage value is found to be 7.6 V. For achieving this collapse voltage and a/t_mto be 16, the membrane (3) should be 5 micron thick and 81 microns in radius.

Finally, as performed in FIG. 14, the optimum amount of delay time between the rising and falling voltage steps is determined. For this particular example, the optimum amount of delay is found to be 22 ns. Using the above method, it is designed a CMUT transducer that would work in the collapse-release regime using 100V pulse in amplitude, 10 ns rise and fall times and 22 ns width. The CMUT cells (2) in the transducer have 5 microns thick membranes (3) with 81 microns radius. The gap (5) height is chosen to be 100 nm and the insulation layer (8) is 200 nm. The membrane (3) and the insulation layer (8) material are assumed to be PECVD silicon nitride. The top and the bottom electrodes (6 and 7) are covering the full surface of the active region. In FIGS. 34 and 35, the simulated ultrasonic pulse is depicted along with its frequency response. It is seen that the designed CMUT transducer using the method described in this document, is capable of transmitting ultrasonic pulses with an amplitude close to 10 MPa, and with a spectrum around 10 MHz. As a second example, it is supposed that there is a need of a pressure pulse signal with an amplitude as high as possible with a spectrum around 25 MHz, using maximum 100V. The transients of applied electrical signal should be as fast as possible for higher output pressures as depicted in FIG. 13. For this specific example let's assume it is managed to achieve rise and fall times of 10 ns for 100V rising and falling voltage steps.

Using the information provided in FIG. 33, it is chosen to use a collapse-release cycle for this case. A full electrode (6 and 7) coverage for both top and bottom electrodes (6 and 7) which would generate the highest possible pressure amplitude for a given voltage as depicted in FIG. 8 is chosen.

The maximum available voltage is limited by the requirements and it is 100V. After that, the insulation layer's (8) thickness is determined by keeping in mind the results of FIG. 9 and 10. The thickness is chosen to be 200 nm, which is believed it would have a dielectric breakdown of more than 100V.

After determining the insulation layer's (8) thickness, the gap (5) height of the membranes (3) for achieving the high pressure output at the desired frequency spectrum is should be chosen. Using FIGS. 21 and 22 it is seen that 100 nm of gap (5) is too large for achieving 25 MHz center frequency at 100V. Choosing smaller gaps (5) if possible increases the electrical forces acting on the membrane (3) and in return increases the pressure output. Therefore for this specific example the gap (5) height is chosen to be 70 nm. Using FIGS. 18 and 22 the a/t_m ratio is chosen about 18. Using FIGS. 18 and 22, it is found out that; in order to achieve a center frequency of 25 MHz and maintain high pressure amplitude; the gamma parameter should be around 10-15. For this particular example, the gamma parameter value is chosen to be 15. Therefore the collapse voltage of the membrane (3) should be 6.7 V. For achieving this collapse voltage and a/t_m to be 18, the membrane (3) should be 2.3 micron thick and 42 microns in radius.

Finally, as performed in FIG. 14, the optimum amount of delay time between the rising and falling voltage steps is determined. For this particular example, the optimum amount of delay is found to be 10 ns. Using the above method it is designed a CMUT transducer that would work in the collapse-release regime using 100V pulse in amplitude, 10 ns rise and fall times and 10 ns width. The CMUT cells (2) in the transducer have 2.3 microns thick membranes (3) with 42 microns radius. The gap (5) height is chosen to be 70 nm and the insulation layer (8) is 200 nm. The membrane (3) and the insulation layer (8) material are assumed to be PECVD silicon nitride. The top and the bottom electrodes (6 and 7) are covering the full surface of the active region. In FIGS. 36 and 37, the simulated ultrasonic pulse is depicted along with its frequency response. It is seen that the designed CMUT transducer using the method described in this document, is capable of transmitting ultrasonic pulses with an amplitude close to 11 MPa, and with a spectrum around 24 MHz.

As a third example, suppose that there is a need of a pressure pulse signal with amplitude as high as possible and with a spectrum around 5 MHz, using maximum 200V. The transients of applied electrical signal should be as fast as possible for higher output pressures as depicted in FIG. 13. For this specific example, suppose it is managed to achieve rise and fall times of 10 ns for 200V rising and falling voltage steps.

Using the information provided in FIG. 33, it is chosen to use a collapse-release cycle for this case. A full electrode (6 and 7) coverage for both top and bottom electrodes (6 and 7) which would generate the highest possible pressure amplitude for a given voltage as depicted in FIG. 8 is chosen. The maximum available voltage is limited to 200V in this example by the system requirements. After that, the insulation layer's (8) thickness is determined by keeping in mind the results of FIG. 9 and 10. The thickness is chosen to be 400 nm, which is believed it would have a dielectric breakdown of more than 200V.

After determining the insulation layer's (8) thickness, the gap (5)height of the membranes (3) for achieving the high pressure output at the desired frequency spectrum is should be chosen. Using FIGS. 21 and 22 it is seen that 200 nm of gap (5) is too large for achieving 5 MHz center frequency. Choosing smaller gaps (5) if possible increases the electrical forces acting on the membrane (3) and in return increases the pressure output. Therefore for this specific example the gap (5) height is chosen as 70 nm. Using FIGS. 15 and 19 the a/t_m ratio is chosen about 17. Using FIGS. 15 and 19, it is found out that; in order to achieve a center frequency of 5 MHz and maintain high pressure amplitude; the gamma parameter should be more than 15. For this particular example, the gamma parameter value is chosen to be 20. Therefore the collapse voltage of the CMUT cell (2) should be 10 V. For achieving this collapse voltage and a/t_m to be 17, the membrane (3) should be 9.3 micron thick and 158 microns in radius.

Finally, as performed in FIG. 14, the optimum amount of delay time between the rising and falling voltage steps is determined. For this particular example, the optimum amount of delay is found to be 48 ns.

Using the above method it is designed a CMUT transducer that would work in the collapse-release regime using 200V pulse in amplitude, 10 ns rise and fall times and 48 ns width. The CMUT cells (2) in the transducer have 9.3 microns thick membranes (3) with 158 microns radius. The gap (5) height is chosen to be 200 nm and the insulation layer (8) is 400 nm. The membrane (3) and the insulation layer (8) material are assumed to be PECVD silicon nitride. The top and the bottom electrodes (6 and 7) are covering the full surface of the active region. In FIGS. 38 and 39, the simulated ultrasonic pulse is depicted along with its frequency response. It is seen that the designed CMUT transducer using the method described in this document, is capable of transmitting ultrasonic pulses with an amplitude close to 10 MPa, and with a spectrum around 5 MHz.

Similar to the examples given above, the disclosed method can be used for designing CMUT cells (2) for achieving the high amplitude pressure pulses for required frequency spectrum and maximum available voltage requirements.

Although the present invention has been described and illustrated in detail, it is clearly understood that the same is by way of illustration and example only and is not to be taken by way of limitation, the spirit and scope of the present invention being limited only by the terms of the appended claims.

Claims

WHAT IS CLAIMED IS:

1. A System (1) for operating capacitive micromachined ultrasonic transducers (CMUT) comprising;

- at least one CMUT cell which is the active acoustical part of a CMUT transducer array or an element or a single cell transducer,

- at least one suspended membrane which is the moving part of the CMUT cell under acoustical or electrical excitation,

- at least one substrate which is the surface on which the fabrication processes are being conducted and the final CMUT transducers are being working,

- at least one gap, which is the vacuum cavity between the substrate and the membrane

and characterized by

- at least one top electrode which is the moving electrode embedded to the moving membrane of the CMUT cell and has a more than 80% coverage of the membrane including the center,

- at least one bottom electrode which is the second electrode of the CMUT cell and which is the stationary electrode embedded to a substrate, and has more than 80% coverage including center of the cell under the moving membrane,

- at least one insulation layer which is an insulating material between the top electrode and the bottom electrode and has a breakdown voltage of Vt,_rk,

- the membrane corresponding to a collapse voltage to collapse to the substrate and to a snapback voltage to resume its suspended position over the gap,

- at least one voltage source for applying a drive voltage signal between the electrodes, with an amplitude greater than 5 times the collapse voltage.

2. The system according to claim 1 is characterized by the voltage source with a drive voltage signal for applying a bias voltage between said electrodes with an amplitude between zero and the collapse voltage, keeping the membrane in suspended position, and thereafter said voltage source for applying a drive voltage between said electrodes, giving a rising voltage step that would bring the voltage level to V_max,, which is greater than 5 times the collapse voltage, to cause the membrane to collapse, and after a time delay, said voltage source for giving a falling voltage step between said electrodes that would bring the final voltage to a value between zero and the snap-back voltage to cause the membrane to return its suspended position.

The system according to claim 1 is characterized by the voltage source with drive voltage signal for applying a bias voltage between said electrodes with an amplitude, V_max, which is greater than 5 times the collapse voltage, keeping membrane in the collapsed position, and thereafter, said voltage source for applying a drive voltage between said electrodes, giving a falling voltage step, which brings the voltage level between zero and the snap-back voltage to cause the membrane to snap-back, and after a time delay said voltage source for giving a rising voltage step between said electrodes that would bring the final voltage to V_max to cause the membrane to return its initial collapsed position.

The system according to claims 2 or 3 is characterized by the voltage source with the time delay between the rising and falling voltage steps of the drive voltage signal to achieve maximum output for the CMUT cell.

The system according to claims 2 to 4 is characterized by the voltage source with the fall time of the falling voltage step of the drive voltage signal to achieve maximum output for the CMUT cell.

The system according to claims 2 to 5 is characterized by the voltage source with the rise time of the rising voltage step of the drive voltage signal to achieve maximum output for the CMUT cell.

The system according to any of the above claims is characterized by the insulation layer's thickness such that V_brk of the insulation layer is more than V_max and less than two times V_max.

8. The system according to any of the above claims is characterized by the CMUT cell which has circular shape.

9. The system according to claim 1 to 7 is characterized by the CMUT cell which has rectangular shape.

10. The system according to claim 1 to 7 is characterized by the CMUT cell which has hexagonal shape.

11. The system according to any of the above claims is characterized by the membrane which is made of silicon nitride.

12. The system according to claims 1 to 10 is characterized by the membrane which is made of silicon.

13. The system according to any of the above claims is characterized by the insulation layer which is made of silicon nitride.

14. The system according to claims 1 to 12 is characterized by the insulation layer which is made of silicon oxide.

15. The system according to any of the above claims is characterized by the substrate which is made of high conductive silicon material.

16. The system according to claims 1 to 14 is characterized by the substrate which is made of glass that is suitable for microfabrication techniques.

17. The system according to any of the above claims is characterized by the top electrode which is buried in the membrane.

18. The system according to claims 1 to 16 is characterized by the top electrode which is at the bottom of the membrane.

19. The system according to claims 1 to 16 is characterized by the top electrode which is at the top of the membrane.

20. The system according to any of the above claims is characterized by the bottom electrode which is a highly conductive metal layer.

21. The system according to any of the above claims is characterized by the insulation layer which is a stationary insulation layer and deposited on top of the stationary bottom electrode.

22. The system according to claims 1 to 20 is characterized by the insulation layer which can be used both at the bottom of the membrane and on top of the bottom electrode.

23. The system according to any of the above claims is characterized by the CMUT cell, which is fabricated using surface micromachining techniques and sacrificial layer process.

24. The system according to any of the above claims is characterized by the CMUT cell which is fabricated using wafer bonding processes.