Patent Application
20150375998
Embodiments of the present disclosure relate to a method for manufacturing a micromechanical system. Further embodiments of the present disclosure relate to a method for manufacturing a pressure sensor. Further embodiments of the present disclosure relate to a micromechanical system. Further embodiments of the present disclosure relate to a method for manufacturing a micromechanical system, and, especially, to the integration of horizontal pressure sensor structures in a CMOS technology.
The term microelectromechanical system (MEMS) or micromechanical system is often used to refer to small integrated devices or systems that combine electrical and mechanical components. When focusing on the micromechanical parts, the term “micromechanical system” may be used to describe small integrated devices or systems which comprises one or more micromechanical elements and possibly, but not necessarily, electrical components and/or electronic components.
Micromechanical systems may be used as, for example, actuators, transducers or sensors, e.g. pressure sensors. Pressure sensors are nowadays mass products in automobile electronics and consumer goods electronics. For many of these applications, systems are used in which the sensor is integrated in an application-specific integrated circuit (ASIC). For example, Infineon Technologies AG offers such a system as a side-airbag sensor.
In particular, the mechanically active elements of a micromechanical system may typically require relatively complex structures, such as recesses, beams, cantilevers, undercuts, cavities etc. Possibly, a relatively high number of manufacturing steps are required. Furthermore, the process used for performing the micromechanical system may need to be compatible with possible subsequent manufacturing steps that are used for creating electrical and/or electronic components, for example.
The present disclosure relates to an integration of e.g. a lamella or a membrane into a 130 nm node, which has been studied only to a minor extend e.g. for a lamella or a membrane integrated within the substrate.
Embodiments of the present disclosure relate to a method for manufacturing a micromechanical system. The method comprises the steps of forming in a front end of line (FEOL) process a transistor in a transistor region after the FEOL process. A structured sacrificial layer is formed at least in a region which is not the transistor region, forming a functional layer which is at least partially covering the structured sacrificial layer, and after the functional layer is formed removing the sacrificial layer in order to create a cavity between the functional layer and a surface, where the sacrificial layer was deposited on.
According to an embodiment, the method may comprise a further act in which a thermal treatment is conducted, wherein the thermal treatment may activate doping atoms of the functional layer and provides a restructuring of the atomic structure of the functional layer. The restructuring of the atomic structure further creates a tension in the functional layer.
Further embodiments of the present disclosure relate to a method for manufacturing a pressure sensor. The method comprises the formation of a sacrificial layer at a surface of the semiconductor substrate, wherein a functional layer covering the sacrificial layer is formed. Furthermore, at least one hole is formed and the sacrificial layer is removed by applying a removing process through the at least one hole to create a cavity. Additionally, a thermal treatment is provided at a temperature between 550° C. and 750° C. and the at least one hole is closed.
Further embodiments of the present disclosure relate to a micromechanical system comprising a semiconductor substrate with at least one transistor in a transistor region and at least one metal layer above the at least one transistor where a lower surface of the at least one metal layer has a first vertical level. Furthermore, the micromechanical system comprises a functional MEMS layer in a MEMS region where the functional layer has an upper surface at a second vertical level which is lower than the first vertical level.
Further embodiments of the present disclosure relate to a method for manufacturing a micromechanical system. The method comprises the step of forming at least one transistor in a transistor region of a substrate. A structured sacrificial layer is provided in a MEMS region of the substrate, where a functional MEMS layer is formed at least partially covering the structured sacrificial layer, and providing etch holes in the functional MEMS layer. Furthermore, the sacrificial layer is removed to create a cavity and a further layer closing the etch holes is formed, wherein the further layer extends above at least a portion of the transistor region and forming at least one metal layer.
Embodiments of the present disclosure are subsequently described in detail using the accompanying figures, in which
Before embodiments of the present disclosure will be described in detail using the accompanying figures, it is to be pointed out that the same or functionally equal elements are given the same reference numbers or similar reference numbers in the figures and that a repeated description for elements provided with the same or similar reference numbers is typically omitted. Hence, descriptions provided for elements having the same or similar reference numbers are mutually exchangeable.
Micromechanical systems, e.g. pressure sensors and other sensors, are mass products. Sometimes the sensor or several sensors are integrated as a system in the form of, for example, an application specific integrated circuit (ASIC). When migrating towards new technology nodes, one of the challenges is to integrate these systems in such a manner that the complexity of, for example, a CMOS (Complementary Metal Oxide Semiconductor) process is not unnecessarily increased and that the existing integration scheme of the CMOS circuit can be preserved to a high extent, preferably with little or minimal development effort.
a-1e show schematic cross sections of a substrate 10 to illustrate a manufacturing process for obtaining a micromechanical system.
The manufacturing process may further comprise a thermal treatment, wherein the thermal treatment activates doping atoms of the functional layer 30 and provides a restructuring of the atomic structure of the functional layer 30. Optionally, the functional layer may cover the sacrificial layer 20 completely. Therefore, a hole may be created in the functional layer 30. The hole may be used to apply a removing process to remove the sacrificial layer 20 to create a cavity between the functional layer 30 and the surface of the semiconductor substrate and wherein a layer is disposed at a surface of the functional layer to close the hole, wherein a height of the layer is below 600 nm. The layer may comprise a nitride or an oxide.
The manufacturing process may further comprise a back end of line (BEOL) process, wherein at least one metal layer is formed in the BEOL, which may be removed at least in a portion of a MEMS region e.g. to structure at least one metal line in the transistor region 27. After the at least one metal line was structured, at least one dielectric layer may be formed in the transistor region and the MEMS region. Furthermore, a thermal treatment may be provided to the functional layer 30 to provide a tension to the functional layer, wherein the temperature of the thermal treatment may be between 550° C. and 750° C.
a-2e show schematic cross sections of a semiconductor substrate 10 to illustrate a manufacturing process for obtaining a pressure sensor 201.
a-3e show schematic cross sections of a substrate 10 to illustrate a manufacturing process for obtaining a micromechanical system.
The functional layer may comprise polycrystalline silicon. Furthermore, the functional layer 30 may optionally comprise a hole, which is closed by a further layer. The height of the layer closing the hole may be below 600 nm. Additionally, the micromechanical system may comprise at least one dielectric layer extending in the MEMS region and the transistor region. Above the dielectric, the micromechanical system may comprise at least one metal layer in the transistor region. Furthermore, a cavity may be formed between the functional MEMS layer and the substrate 10.
a-5d show schematic cross sections of a processed substrate 10 and related structures to illustrate an example manufacturing process for obtaining a micromechanical system. The manufacturing process comprises the formation in a front end of line (FEOL) process of a transistor 25 in a transistor region 27, wherein, after the FEOL process, a structured sacrificial layer 20 is formed at least in a region which is not the transistor region 27. A functional layer 30 is formed which is at least partially covering the structured sacrificial layer 20, and, after the functional layer 30 is formed, the sacrificial layer 20 is removed in order to create a cavity 35 between the functional layer 30 and a surface 15, having the sacrificial layer deposited thereon.
In general, the FEOL-process typically comprises the formation of the semiconductor substrate which may comprise the provision, planarization and cleaning of the wafer, depositing a shallow trench isolation (STI) and forming a well. Additionally, source, drain and gate terminals are formed which may build electrical or electronic components such as transistors. In other words, the FEOL process comprises the preparation of the semiconductor substrate whereupon metallization structures, which are separated by an insulation material, may be applied in a further back end of line (BEOL) process. The process described in the present disclosure may be applied between the FEOL and the BEOL process.
To be more specific,
After the FEOL process, a sacrificial layer 20 is deposited by means of, for example, a chemical vapor deposition (CVD) process, an ion deposition process, an ion-assisted sputtering, a sputtering process, a cathodic vacuum arc process, a plasma deposition process, and/or a pulsed laser deposition at the surface 15 of the processed semiconductor substrate 10. The sacrificial layer 20 may comprise carbon materials. Furthermore, the sacrificial layer 20 may be structured, e.g. by a lithography and etching process, or it may be deposited as a structured sacrificial layer. A lateral boundary may be defined to laterally delimit the sacrificial layer so that it covers a certain portion of the surface of the substrate, wherein a remaining portion of the surface of the substrate is exposed or at least not covered by the sacrificial layer 20. The sacrificial layer may also be structured to have a lateral shape with at least one vertex or corner. The structure of the sacrificial layer defines the dimension at an accuracy of, e.g. 20 nm, and a height at an accuracy of, e.g. less than 6%, e.g. ±2 . . . 3%, of a cavity, which remains after the sacrificial layer is removed in a further processing step.
b shows the processed semiconductor substrate 10, wherein a functional layer 30 is deposited on the surface 15 of the processed semiconductor substrate 10 such that the sacrificial layer 20 is covered. The functional layer 30 may be deposited such that the sacrificial layer 20 is covered at least partially, e.g. by using a further lithography step to structure the functional layer 30, or such that the functional layer 30 covers the sacrificial layer 20 completely, as it is shown in
The functional layer 30 may comprise an in situ p-doped amorphous silicon, e.g. comprising boron in the dopant, in order to derive required or desired electrical properties and/or mechanical properties, such as a higher mechanical stiffness or a better electrical or thermal conductivity. The functional layer may be deposited using a polylayer in situ deposition and structuring. The process requires a comparably low temperature of, e.g. between 300° C. and 600° C., or between 400° C. and 500° C., or about 440° C., which may preserve the electrical properties of the processed semiconductor substrate.
c shows the processed semiconductor substrate 10, wherein the sacrificial layer 20 is removed by a removing process, such that a cavity 35 between the surface of the processed semiconductor substrate 15 and the functional layer 30 is obtained. The removal process may comprise at least one of a thermal decomposition process, a dry etching process, and an ashing process. In case the sacrificial layer 20 is covered completely with the functional layer 30 (see
If the functional layer 30 covers the sacrificial layer 20 only partially, the removing process can be applied through those areas where the sacrificial layer 20 is not covered by the functional layer 30.
The removal of the sacrificial layer 20 may be supported by the use of a material that exhibits an extreme selectivity regarding the etch rate to the etch velocities of the surrounding materials. In this manner, the complete sacrificial layer 20 may be removed even from remote points of the functional layer 30. Furthermore, the removal may be obtained by a dry etching process such that problems such as sticking during the process are not to be expected. The resulting cavity may have a lateral range of e.g. 20μm at an accuracy of e.g. 20 nm and a height of e.g. less than 500 nm, less than 300 nm, or less than 100 nm, at an accuracy of the height of e.g. less than 15%, less 10%, or less than 5%.
A further act may be the application of a thermal treatment to the functional layer 30. The thermal treatment may cause a restructuring of the atomic structure of the functional layer, i.e. the functional layer may be tensed and/or healed. Further, the doping of the functional layer may be activated. The tension, e.g. an (intrinsic) tensile stress, may be obtained by e.g. a heat induced transformation of amorphous silicon to polycrystalline silicon in the functional layer. This combination of the activation of the doping and the restructuring of the atomic structure in a single thermal treatment is a further step in this disclosure. Furthermore, the comparably low temperature in an optimal range of, e.g. 450° C. to 850° C., or 550° C. to 750° C., and the short duration of the thermal treatment time of, e.g. less than 90 minutes, less than 60 minutes, or less than 30 minutes, limits the impact on the processed semiconductor substrate 10 after the FEOL-process. Thus, the transistor(s) 25 and optionally build other electronic components preserve their specifications. Otherwise, they would need to be redefined in an elaborate process.
d shows the micromechanical system where an optionally deposited layer 45 may cover the surface 15 of the processed semiconductor substrate 10, the functional layer 30, and the transistor 25. If the hole 75 has been created previously, the layer 45 may be arranged inside the hole such that the cavity 35 is sealed. At the same time, a portion of the cavity 35 that is beneath the at least one hole 75 might also be filled with the same filling material 77. This portion of the cavity is typically filled due to a side effect of filling the at least one hole. As mentioned above, the at least one hole 75 is adjacent to, or in an extension of, the margin region of the sacrificial layer 20 and hence also adjacent (typically in a direction orthogonal to the surface 15 of the substrate 10) to a margin region of the cavity 35. Nearby side walls, e.g. a part of the functional layer 30, and the bottom bounding the cavity 35 may lead to a locally confined deposition of the filling material 77 within the cavity (e.g. due to adhesion effects) so that the portion of the cavity 35 that is filled makes up only a small fraction of the entire cavity. Moreover, the filled cavity portion is within a margin portion of the cavity 35, i.e. close to a lateral boundary of the cavity, so that a clear width of the cavity is reduced only to a minor or negligible extent by the filled cavity portion. The filling material 77 may be a material forming or comprised by the layer 45, which may be e.g. an inter layer dielectric (ILD) material of an ILD layer.
The margin region of the sacrificial layer 20 may be regarded as a region close to a lateral circumference of the sacrificial layer, compared to an overall lateral extension of the sacrificial layer 20. The width of the margin region may depend on the geometric structure and/or the micromechanical properties of the functional layer 30. For example, the margin region may be relatively wide in an area near the corners of the sacrificial layer 20, and more narrow near the centers of the edges. Alternatively or additionally, the lateral cross sectional area of the at least one hole 75 may be defined as being at most 5% of the lateral cross sectional area of the cavity 35, optionally, preferably at most 4% or 3% or 2% or 1%. From these optional definitions it becomes clear that a portion of the functional layer that is adjacent to a central portion of the sacrificial layer (and an eventually created cavity after removal of the sacrificial layer) is free of holes, as these are (exclusively) located in the margin region. According to some embodiments the margin region may extend from the lateral boundary of the sacrificial layer in a direction perpendicular to the lateral boundary by at most 30% of a largest lateral extension of the sacrificial layer 20.
Arranging opening points, e.g. holes 75, in the margin/corner region allows for a high degree of freedom when closing the openings or hole 75. Since the corners only play a minor role with respect to the membrane's (functional layer's) mechanical properties and displacement under forces acting in orthogonal direction to the membrane (functional layer), maintaining the cavity 35 (immediately) under the further layer 45 closing the hole 75 is not absolutely mandatory (i.e., maintaining the cavity portion directly under the filled hole and in the vicinity).
Furthermore, it may be also possible to remain the hole 75 opened, e.g. to allow a pressure compensation. Additionally, it may be also possible to create a hole 75 in the functional layer 30 which partially covers the sacrificial layer 20, e.g. to optimize the removing process.
In order not to influence further processing acts, e.g. a polishing act, the layer may comprise a maximum height that is, e.g., less than 1500 nm, less than 1000 nm, or less than 600 nm. The height may be the distance between the surface 15 of the processed semiconductor substrate 10 and an exposed surface of the layer 45, perpendicular to the surface of the processed semiconductor substrate 10. The layer 45 may comprise one of a nitride or an oxide.
The deposition of the layer 45 is processed in a high vacuum environment, comprising a pressure of, e.g. less than 100 mTorr, less than 50 mTorr, or less than 10 mTorr. The vacuum may support the cavity evacuation from e.g. residuals of the etching process or a gas of a dry etching process using high density plasma. In other words, it may be a joint evacuation, closure and ILD deposition obtained. The layer 45 may be a first step of a BEOL process comprising e.g. further CMOS processes such as a surface polishing step.
The BEOL-process may comprise building a horizontal stack of metallization structures, e.g. metal layers or metal lines, and intermediate dielectric layers. The stack is formed or deposited to form a wiring or interconnections of the micromechanical system. Typically, the metal layers are deposited and structured, e.g. in a lithography process, to form a metal line between different transistors 25, wherein the metal line may form an electrical connection. A dielectric layer isolates the connections and builds a basis for a further metal layer, which is again structured. Connections between different metallization layers may be e.g. provided by vias through the dielectric layers. The steps of depositing a dielectric layer, depositing a metallization layer, and structuring the metallization layer to form at least one metal line may be repeated until the whole wiring is completed. Additionally, further layers may be deposited in the stack forming intermediate layers, for example, to form an etch stop for an etching agent. Furthermore, at least in a part of the MEMS-region, all layers of the BEOL-process, i.e. every layer that is deposited after the functional layer, is removed to provide access to the functional layer. The removed part of the layers forms a trench in the stack. Nonetheless, there are no further high temperature steps necessary which are able to harm the processed semiconductor substrate, especially the doping of the processed semiconductor substrate, or the tension and/or the doping of the functional layer.
In the context of technologies with structure widths of 130 nm and below, for example the topology of front-end-of-line structures plays a relatively important role. Embodiments allow that a total thickness does not exceed a few hundred nanometers, in order not to impair a BPSG (borophosphorsilicate glass) polishing step which is often present. Furthermore, according to the embodiments, a reduction for as few as possible intermediate steps and the sharing or joint use of existing processes is achieved to facilitate a successful integration, i.e. the prerequisites for a successful integration are fulfilled.
With respect to pressure sensors, one option is provided by integrated pressure sensor structures that use a polycrystalline silicon blade or a membrane which is formed on a substrate. The used process employs 5 to 6 mask layers and an elaborate process for removing a sacrificial layer between the blade and the substrate. In this form, the present process is hardly integratable in a CMOS process. The present disclosure only uses 2 to 3 mask layers and a CMOS compatible removing of the sacrificial layer and is therefore much better integratable in a CMOS process, even in connection with an integration into a 130 nm node.
Furthermore, a doped semiconductor, e.g. the functional layer, typically requires an activation, e.g. if an ion implantation technique is used. The dopant first remains on an interstitial of the crystal lattice of the substrate, which comprises e.g. silicon, where the dopant has no effect. Additionally, the ion implantation destroys parts of the crystal lattice due to collisions of the dopant shot into the lattice and molecules of the crystal lattice, e.g. silicon molecules. A thermal treatment, e.g. an annealing, of the doped semiconductor activates the doping, i.e. molecules of the dopant are integrated in the crystal lattice, and the processed semiconductor substrate is healed by replacing missing molecules, both due to an increased thermal motion of the molecules. Besides these positive effects, a diffusion of the dopant takes place. The diffusion increases with an increased duration and an increased temperature of the thermal treatment. Therefore, embodiments of this disclosure achieve a limitation of the thermal treatment.
To this end, if there is a hole 75 created in the functional layer 30, which is e.g. a membrane, the hole(s) 75 may be performed at the corner(s) or the margin of the sacrificial material or the cavity 35. Hence, the lithography process for structuring the polycrystalline blade can be used, in a joint manner, for hole etching. Furthermore, it is possible to use a process for closing, filling, or plugging, which does not necessarily has to preserve a portion of the cavity 35 (directly) under the hole 75 within the blade. The process might be included in the deposition of the layer 45, which may be e.g. an ILD layer.
The functional layer 30 may be formed as a polycrystalline silicon blade or a membrane and may build a pressure sensor. The variant proposed integrates the pressure lamella once the CMOS structures have been completed, and reduces the number of structuring lithography processes. Depending on the process management, it is also possible to simultaneously use the sealing oxide as an ILD prior to contact hole structuring. In this manner, maximum utilization of existing processes within the CMOS process is achieved.
Opening of the membrane is performed, for this purpose, exclusively on the corners or sides, so that, for one thing, the lithography process may be used for structuring the polylamella for hole etching; in addition, it becomes possible to utilize, for the closure or sealing, a process which does not necessarily need to maintain the cavity 35 below the hole 75 within the lamella.
By means of suitable temperature treatment, the lamella is modified such that in addition to activation of the dopants, tensile stress is also achieved. All of the layers supported on the sensor lamella are removed later on by means of a dry etching process and in one implementation variation replaced by thin passivation. Further specific processes are the dry evacuation of the cavity (e.g. etching of the sacrificial layer), activation of the dopants and the application of a tensile stress to the lamella, and the combination of closure or sealing of the evacuation holes and ILD deposition.
According to embodiments, a suitable selection of the integration concept for a pressure sensor lamella is achievable. This includes insertion of the lamella once essential parts of the transistor structures of the surrounding CMOS circuit have been completed (following FEOL processing), temperature treatment with suitable parameters in a manner which does not impair the CMOS circuit, and simultaneous evacuation of the cavity and deposition of part of the bottommost interlayer dielectric, in which the bottommost contact level will be structured at a later point in time. Thus, existing process sequences may be used in the CMOS process (only 2 to 3 additional mask levels for integrating a pressure sensor).
The at least one transistor 25 may be configured to provide, e.g. an electrical circuitry, for example to transduce or process a variation of physical properties of the functional layer. The at least one metal layer 65 above the at least one transistor 25 is configured to provide an electrical contact to at least one transistor 25. The functional layer 30 in the MEMS region 60 is, for example, a functional part in a sensor, wherein the functional layer 30 is configured to change its physical properties, e.g. due to different ambient pressures (acting on the functional layer 30). The physical properties of the functional layer may be, e.g., the conductivity of the functional layer. The layer 45 and optionally more/other layers, e.g. horizontally stacked ILD and metal layers/metallization structures/metal lines 70, covering the functional layer 30, may extend above at least a portion of the transistor region 27 during the manufacturing process. Nonetheless, the layers are removed in further steps of the manufacturing process at least in a part of the MEMS area 60, forming a trench 85 to provide access to the functional layer 30, which is, e.g. a sensory part of the micromechanical system.
The well 50 is located adjacent to the surface 15 of the semiconductor 10. Typically, the well 50 comprises the same doping type as the doped substrate 200, however at a higher doping concentration resulting in a higher electrical conductivity of the well 50 compared to the doped substrate 200. For example, both the doped substrate 200 and the well 50 may be p-doped, and the doping level of the well 50 may be 2 to 5 orders of magnitude higher than the doping level of the doped substrate 200. The high doping concentration of the well 50 may be created by means of an ion implantation process or a diffusion process, followed by a thermal treatment such as a furnace anneal or a rapid thermal anneal (RTA). The well 50 may be used as a (counter) electrode in a manufactured micromechanical system or micromechanical pressure sensor.
The functional layer may eventually form one of a membrane, a beam, a lever, a cantilever, etc. of the micromechanical system once the micromechanical system is build. The functional layer thus becomes a mechanical transducing or mechanically active element. Because of the cavity that has been formed adjacent to the functional layer, at least a portion of the functional layer is able to move in response to a mechanical force, a pressure, an acceleration, an electrical force, and other physical influences.
The previously discussed embodiments allow the efficient implementation of pressure sensors. The singulation of the sensors may be performed by any appropriate technique, such as by cutting, sawing, etching or stealth dicing. In stealth dicing, a laser destroys the crystal-structure in a predetermined depth of the substrate, such that the substrate may be singulated (for example by breaking the substrate) such that the cavities of the substrate can be reliably opened and ventilated.
Although some aspects have been described in the context of an apparatus, it is clear that these aspects also represent a description of the corresponding method, where a block or device corresponds to a method act or a feature of a method act. Analogously, aspects described in the context of a method step also represent a description of a corresponding block or item or feature of a corresponding apparatus. Some or all of the method acts may be executed by (or using) a hardware apparatus, like for example, a microprocessor, a programmable computer or an electronic circuit. In some embodiments, some one or more of the most important method acts may be executed by such an apparatus.
In the foregoing Detailed Description, it can be seen that various features are grouped together in embodiments for the purpose of streamlining the disclosure. This method of disclosure is not to be interpreted as reflecting an intention that the claimed embodiments require more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive subject matter may lie in less than all features of a single disclosed embodiment. Thus the following claims are hereby incorporated into the Detailed Description, where each claim may stand on its own as a separate embodiment. While each claim may stand on its own as a separate embodiment, it is to be noted that—although a dependent claim may refer in the claims to a specific combination with one or more other claims—other embodiments may also include a combination of the dependent claim with the subject matter of each other dependent claim or a combination of each feature with other dependent or independent claims. Such combinations are proposed herein unless it is stated that a specific combination is not intended. Furthermore, it is intended to include also features of a claim to any other independent claim even if this claim is not directly made dependent to the independent claim.
It is further to be noted that methods disclosed in the specification or in the claims may be implemented by a device having means for performing each of the respective steps of these methods.
Furthermore, in some embodiments a single act may include or may be broken into multiple sub acts. Such sub acts may be included and part of the disclosure of this single act unless explicitly excluded.
The above described embodiments are merely illustrative for the principles of the present disclosure. It is understood that modifications and variations of the arrangements and the details described herein will be apparent to others skilled in the art. It is the intent, therefore, to be limited only by the scope of the impending patent claims and not by the specific details presented by way of description and explanation of the embodiments herein.
