Method for handling MEMS chips

Abstract

The invention relates to a method for handling MEMS chips (120) having MEMS structures (122), in particular sensitive MEMS structures (122), which are arranged on a top surface (174) of the MEMS chips (120). At least one handling surface (170, 172), which is preferably formed laterally on the MEMS chips (120) and is connected to the MEMS chips (120) via connecting bridges (208), is provided (300). The at least one handling surface (170, 172) is produced during the processing of a wafer (100) on one or more chip sides (178, 180) or chip side sections (182) of the MEMS chip (120). At least one protective wafer (194) is attached (310) to the laterally formed handling surfaces (170, 172) which protect the sensitive MEMS structures (122) from damage while process steps (312-370) are being carried out. The laterally extending handling surfaces (170, 172) are then removed (380) in a substantially force-free manner after the MEMS chips (120) have been processed. The invention also relates to the use of the method to produce and place MEMS chips (120) on a substrate (160) while maintaining minimised distances (212) between them.

Description

[0001] Description

[0002] title

[0003] Methods for handling MEMS chips

[0004] Technical area

[0005] The invention relates to a method for handling MEMS chips with MEMS structures, in particular sensitive MEMS structures arranged on a top side of the MEMS chips. Furthermore, the invention relates to a MEMS chip with at least one handling surface and to the use of the method for handling and manufacturing MEMS chips.

[0006] State of the art

[0007] WO 2017 / 129171 A1 discloses a method and apparatus for separating a microchip from a wafer and applying the microchip to a substrate. During the detachment process, the microchip is snapped onto the free end of a tip. It adheres to the tip by adhesive force during transport to the substrate. Until the microchip is detachable, it is held in place by webs known as retaining tabs. These webs, which serve as retaining tabs, are broken during detachment.

[0008] DE 10 2018 214 017 A1 relates to the production of a thin film with a microsystem, wherein the microsystem is produced on the thin film and peeled off together with the thin film. Among other things, it is proposed to form webs between an outer region and an inner region of the thin film, with the microsystem located on the inner region of the thin film. When the inner region is peeled off, the webs can break. A support structure can be formed which extends over the inner region and at least partially over the webs and which is peeled off together with the inner region. It is further disclosed to design the webs with tapered regions in order to define predetermined breaking points.

[0009] US 2008 / 0044985 A1 discloses the use of sacrificial support parts on workpieces with micro features, such as semiconductor wafers, wherein the workpiece comprises a substrate and microelectrical dies. Furthermore, the workpiece has a sacrificial support part which is attached over the dies and serves for protection and transport. It is disclosed to carry out a singulation of such a workpiece, wherein the singulation also relates to the sacrificial support part. A singulated die is applied to a carrier substrate, for example an interposer substrate, using the associated part of the sacrificial support part. Subsequently, the sacrificial support part is removed. Using such an approach, a device with micro features, for example a microelectronic or a micromechanical device, can be produced.

[0010] WO 2017 / 129171 A1 and DE 10 2018 214 017 A1 describe various possibilities for singulating wafers and then transporting the singulated chips. The use of webs is disclosed, with DE 10 2018 214 017 A1 also proposing the use of specially shaped webs with tapered sections.

[0011] US 2008 / 0044985 A1 discloses the general idea of ​​using special elements for transporting chips and their subsequent removal.

[0012] For MEMS components, which have sensitive, movable elements, such as mirror elements, located across the entire surface, especially on the front of the component, the separation and subsequent handling of the components for further processing is difficult. Handling of the separated component can only be achieved contactlessly, from the back, the side, or on additional auxiliary surfaces that serve no other function.

[0013] The use of contactless handling techniques, such as Bernoulli handling or acoustic levitation, is often impossible due to the very weak coupling to the component. Such techniques also have the disadvantage that the moving, sensitive MEMS structures can be damaged by the gas flow during Bernoulli handling or by the acoustic vibrations themselves during acoustic levitation. Acoustic levitation also has limitations due to the geometry of the immediate environment influencing the required sound field.

[0014] Handling from the back is generally not possible when assembling a component onto a substrate, for example when soldering it on.

[0015] If auxiliary surfaces are provided outside the MEMS structures for attaching protective caps and for handling, these reduce the achievable (optical) fill factor. Without such a frame, the individual components for assembly in an array can only be handled on two of the six chip side surfaces.

[0016] Disclosure of the invention

[0017] A method for handling MEMS chips is proposed, with MEMS structures, in particular sensitive MEMS structures, which are arranged on an upper side of the MEMS chips, wherein the following method steps are carried out: a) providing at least one handling surface, which is preferably formed laterally on the MEMS chip and is connected to the MEMS chip via connecting webs, b) fastening at least one protective wafer on a front side of the wafer, whereby the sensitive MEMS structures are protected from damage during the execution of process steps, c) exposing and singulating MEMS chips during processing of the wafer and producing handling surfaces running laterally along MEMS chips and d) force-free detachment of the handling surfaces after processing and placement of the MEMS chips.

[0018] Advantageously, in the method proposed by the invention, the MEMS component or a MEMS chip can be handled on the laterally formed handling surfaces and applied to a substrate. A fixed mechanical connection can be released by a laser process, whereby MEMS components in the form of MEMS chips can be separated from a wafer assembly and / or the handling surfaces can be detached from MEMS chips.

[0019] In an advantageous embodiment of the method proposed according to the invention, the MEMS chips comprise sensitive MEMS structures for microelectromechanical systems, wherein the MEMS structures are exposed before a wafer is singulated.

[0020] In the method proposed according to the invention, the MEMS structures, in particular the sensitive MEMS structures, comprise or are those which are suitable for one or more MEMS components, such as MEMS sensors and MEMS actuators, for example MEMS inertial sensors, MEMS pressure sensors, MEMS microphones, MEMS micromirrors and / or MEMS resonators.

[0021] In the method proposed by the invention, the MEMS chips of the wafer are advantageously electrically connected at least partially to one or more components before the wafer is singulated. These components are advantageously ASICs that can be contacted via a backside metallization.

[0022] The method proposed according to the invention is further characterized in that, according to method step a), a pre-structured silicon wafer with solderable backside metallization is provided as the wafer.

[0023] In the method proposed according to the invention, it is advantageously provided that, according to method step b), at least one protective wafer is fastened to a front side of the wafer by means of temporary bonding, which protective wafer comprises a cavity and / or at least one opening and / or an optical window.

[0024] The method proposed by the invention further advantageously provides that, according to method step c), silicon structures are dissolved out by etching in SFe and / or XeF2, and externally exposed protective oxide is removed by HF gas-phase etching. The method proposed by the invention, according to method step c), allows the MEMS chips in the wafer to be provided with a circumferential trench on the side during the release process, thus exposing them.

[0025] The method proposed according to the invention is further characterized in that, after performing method step c), the MEMS chips are locally connected to the wafer via connecting bridges, and a protective layer on the solderable backside metallization of the wafer can be removed. The remaining connecting bridges formed in the wafer material enable the creation of a wafer assembly from which the MEMS chips, exposed by circumferential trenches around the individual MEMS chips, can be removed using a pick-and-place tool.

[0026] The method proposed by the invention further provides that the wafer resting on the front side protected by at least one protective wafer is electrically contacted on its back side, and ASICs are applied to the back side, electrically contacted, and measured. Thus, several work steps can be performed in the wafer position where the back side is freely accessible for processing operations.

[0027] In the method proposed by the invention, the temporary bonds of at least one protective wafer are released and the connecting webs separated by means of a full-surface or partial laser exposure. This measure advantageously allows for processing of the wafer material, so that subsequent processing steps can be carried out more easily.

[0028] In the method proposed by the invention, according to method step c), isolated MEMS chips are temporarily held in the wafer assembly by the handling surfaces and can be removed from the wafer assembly using a pick-and-place tool. Advantageously, the method proposed by the invention provides for placing the isolated MEMS chips on the substrate and then performing a substantially force-free detachment of the handling surface from the MEMS chips.

[0029] Furthermore, the method proposed according to the invention is characterized in that a removal of MEMS chips from the wafer assembly and a mounting on the substrate is repeated, so that a number of MEMS chips are assembled at a minimized distance from one another to form a MEMS array.

[0030] Furthermore, the invention relates to a MEMS chip with at least one handling surface, manufactured according to the method described above, wherein bridge webs between a chip edge of the MEMS chip and the handling surface(s) are formed essentially by a first triangle and a second triangle, which have an overlap zone in which either a structuring or weakening zone or a cross-sectional taper of the silicon material of the wafer is present. This advantageously allows for rapid separation of the bridge webs without requiring excessive energy input.

[0031] Finally, the invention relates to the use of the method for producing and handling MEMS chips and their placement on a substrate while maintaining minimized distances between the MEMS chips.

[0032] Advantages of the invention

[0033] The solution proposed by the invention provides a separation, protection, and handling concept in which additional, essentially L-shaped handling surfaces are created on the sides of the MEMS components in the form of MEMS chips. These laterally arranged handling surfaces enable the attachment of caps and socket structures as protective wafers, which serve as protection against damage during individual processing steps, such as sacrificial layer etching, electrical and optical testing during the rear-side assembly of electronic components, such as ASICs, and during separation. With these additional lateral handling surfaces, the MEMS chips can then be assembled together with the ASICs to form a planar array with a maximum optical fill level.Advantageously, the handling surfaces, designed as laterally extending, L-shaped surfaces, can be removed after the assembly of the MEMS chips.

[0034] The method proposed according to the invention defines a manufacturing process for singulated and electrically tested MEMS chips with an electrically contacted, in particular soldered, ASIC on the back of a wafer, which enables such handling. In particular, the MEMS chips can be removed from the wafer assembly after exposure, and the separation of the essentially lateral handling surfaces from the MEMS component can be performed after the assembly process.

[0035] The method proposed according to the invention allows for the effective protection and handling of MEMS chips, for example, designed as tiltable mirror elements. It enables the MEMS structures to be exposed through etching processes, after which they become movable in the wafer, the soldering of electronic components in the form of ASICs to the back of the MEMS, and their electrical or optical measurement with and without ASICs while maintaining the MEMS chip position in the wafer. Furthermore, it is advantageously possible to separate the MEMS chips from the wafer assembly, which is essentially made of pre-structured silicon (singling), and to mount the MEMS chips with an electrically contacted ASIC on a substrate to form an array with maximum fill level.

[0036] The handling surfaces, which essentially run laterally along the chip edges of the MEMS chips, are preferably attached to only one or two sides, or even to side sections of the sides of the MEMS components. The stability of the connection between the handling surfaces and the MEMS chips can be adjusted via the length, width, and thickness of the mechanical connecting bridges. After the MEMS component has been mounted on a substrate, such a mechanical connection in the form of connecting bridges with targeted structuring zones, weakening zones, or cross-sectional reduction zones can be released in such a way that no elements protruding beyond the chip edge hinder the assembly of an adjacent MEMS chip, which is a prerequisite for a maximum achievable fill factor.

[0037] The number, position, and geometry of the connecting bridges can be adapted to the substrate depending on the required load during electrical testing or during assembly of the ASIC and MEMS chip. The mechanical strength of the connecting bridges is essentially determined by their thickness, width, and length. The larger these dimensions, the more resilient the connection is from a mechanical perspective, but the more difficult it is to separate these connections using a laser. In the method proposed by the invention, the connecting bridges or bridging bridges are separated by introducing, for example, a local effect in the region of the narrowest point, preferably within the overlap area. A mechanical stress gradient is created by means of local laser energy coupling and adjacent local cooling.The defect preferentially propagates as a crack in this mechanical stress field and can be guided through the stress field where mechanical stress is present (thermal laser suppression). Furthermore, separation is possible by laser ablation of the wafer material in the area of ​​the narrowest points of the connecting bridges, as well as by melting the material in the area of ​​the narrowest points of the connecting bridges.

[0038] The solution proposed by the invention enables high mechanical strength to be achieved between the individual MEMS chips and their handling surfaces. This relatively high mechanical strength allows the application of greater forces, which represents a decisive advantage for backside assembly, for example with an electronic component such as an ASIC, as well as for electrical measurement and precise component assembly on the substrate. Furthermore, the solution proposed by the invention enables the individual MEMS chips to be exposed to the outside in such a way that, when the individual MEMS chips to be exposed are covered with at least one protective wafer, the top sides of the MEMS chips, with the MEMS structures extending thereon, in particular over the entire surface, can be held within the wafer assembly via the at least one protective wafer.The method proposed according to the invention enables handling of MEMS chips with full-surface MEMS structures. The handling of such chips is made possible by the method proposed according to the invention, in particular, by the handling surfaces, which are essentially L-shaped and run along the sides of the chip edges. The method proposed according to the invention also allows the MEMS chips, with the exception of the connecting webs, to be removed from the wafer assembly all the way around by forming trenches during the release process. Even after the connecting webs have been separated, the MEMS chips can remain in a wafer assembly, which, however, is formed by the protective wafers assigned to the respective MEMS chips.

[0039] Short description of the drawings

[0040] Embodiments of the invention are explained in more detail with reference to the drawings and the following description.

[0041] They show:

[0042] Figure 1 is a schematic plan view of an assembly of several MEMS chips to form a MEMS array,

[0043] Figures 2.1 - 2.8 show schematic cross-sectional views of a structure and a process sequence for the production and release of the MEMS chips for their electrical / optical measurement, the application of ASICs, the removal of the MEMS chips from a wafer assembly for the mounting of the MEMS chips on a substrate,

[0044] Figures 3.1 and 3.2 show a representation of a MEMS chip with a connected handling surface and a representation of a MEMS chip with a separated handling surface (L-Rim),

[0045] Figure 4 shows a plan view of the chip arrangement within a wafer assembly, wherein the MEMS chips are held by connecting bridges, Figure 5 shows the structure of an array of MEMS chips with the illustration of the bridge bridges between the handling surface and the MEMS chip edge,

[0046] Figure 6 is a plan view of the chip arrangement in the wafer assembly and an enlarged view of a connecting bridge,

[0047] Figure 6.1 a cross-section before separation,

[0048] Figure 7 shows a detailed view of the connecting bridge with structuring in plan view,

[0049] Figure 7.1 the connecting web shown in Figure 7 in cross section,

[0050] Figure 8 shows a detailed view of a bridge web after laser separation,

[0051] Figure 8.1 a cross-sectional view of Figure 8

[0052] Figure 8.2 a side view of a bridge web with a cross-sectional reduction in the area of ​​the narrowest point,

[0053] Figure 9 shows another variant of the bridge webs and

[0054] Figure 10 is a schematic representation of the processing steps of the method proposed according to the invention.

[0055] Embodiments of the invention

[0056] In the following description of the embodiments of the invention, identical or similar elements are designated by the same reference numerals, whereby a repeated description of these elements is omitted in individual cases. The figures only schematically illustrate the subject matter of the invention.

[0057] Figure 1 shows an illustration of an assembly of multiple MEMS chips 120 to form a MEMS array 110. As can be seen from Figure 1, handling surfaces 170, 172 are located on two possible MEMS chips 120, namely along the first chip side 178 and along a second chip side 180. Protective wafers 194, such as those shown in Figures 2.2, 2.3, and 2.4, can be attached for handling on the handling surfaces 170, 172, shown here in plan view, which are essentially L-shaped and have a first leg 184 and a second leg 186 arranged perpendicular to the first leg. After the MEMS chip 120 has been mounted in its position on the MEMS array 110, the handling surfaces 170, 172 attached to the chip sides 178 and 180 are removed. The detached edge structure 188 is shown in dashed form in Figure 1.

[0058] In this way, the individual MEMS chips 120 can be mounted with a minimized distance 212, as will be described below, on a substrate 160, as shown schematically in plan view in Figure 1, whereby a very high optical fill factor can be achieved.

[0059] The method proposed according to the invention, which is only roughly outlined with reference to Figure 1, enables placement of the MEMS chip 120, which comprises very sensitive MEMS structures 122 on its upper side 174, by means of a pick-and-place tool (not shown in detail here) in the placement direction 176, for example diagonally to the upper side of the substrate 160. Reference numeral 182 designates side sections of the first chip side 178 and the second chip side 180, along which handling surfaces 170, 172 could also be formed, alternatively to the representation of the continuously L-shaped legs 184, 186 according to the schematic view in Figure 1.

[0060] In the sequence of figures 2.1 to 2.9, a process sequence for the production and release of the MEMS chips 120, their electrical or optical measurement, application of ASICs 204 and the removal of the MEMS chips 120 with handling surfaces 170, 172 together with ASIC 204 from a wafer assembly 214 and mounting of the MEMS chips 120 on a substrate 160 is shown in more detail.

[0061] The illustration in Figure 2.1 shows that MEMS chips 120 are embedded within a wafer 100. Wafer 100 is, for example, a silicon wafer pre-structured using EPYC technology, which has a backside metallization 190 provided with a protective layer. Wafer 100, preferably formed as a pre-structured silicon wafer, has a front side 192 and a back side 206. Each of the MEMS chips 120, depicted here as MEMS components, comprises MEMS structures 122 that are directed to the front side 192 of the pre-structured silicon wafer 100.

[0062] Figure 2.2 shows that one or more protective wafers 194, also referred to as cap wafers, are applied to the front side 192 of the wafer 100, which is preferably a pre-structured silicon wafer, using a temporary bonding process. The at least one protective wafer 194 has a cavity and / or openings 196 for etching media in the area of ​​the MEMS structures 122, which are required for exposure. Alternatively, it is possible to design the at least one protective wafer 194 such that it also has a full-surface opening or an optical window 198 with smaller openings (not shown in detail here). The MEMS structures 122 can be, for example, MEMS micromirrors, which are extremely sensitive to damage.

[0063] Figure 2.2 shows, analogously to the illustration in Figure 2.1, that the wafer 100, preferably a pre-structured silicon wafer, comprises the front side 192 and the back side 206. On the back side 206, backside metallizations 190 are applied to the wafer 100, which are covered by a protective layer.

[0064] Figure 2.3 shows that when performing the isolation of the individual MEMS chips 120 from the wafer 100, outwardly exposed silicon structures are first removed using an etching process, for example, RI E-etching, in SFe and / or XeF2. The outwardly exposed protective oxide of the structures is then removed using RF gas-phase etching. During the isolation, previously fixed structures become movable and are thus extremely sensitive to vibrations and contact. During the isolation process, which is shown in Figure 2.3, the individual MEMS chips 120 embedded in the wafer 100 are exposed laterally, except for a few locations, by means of a circumferential trench 200. They are then only locally connected to the material of the wafer 100 via connecting webs 208 and thus retain their position within the wafer assembly 214 shown in Figure 2.3. During the isolation process as shown in Figure 2.3, the protective layer provided on the solderable backside metallization 190 on the backside 206 of the wafer 100 is also removed.

[0065] Figure 2.3 shows that the front side 192 of the wafer 100 is still protected by the protective wafers 194 arranged side by side and attached to the silicon material of the wafer 100 by means of temporary bonds. After the exposure and the creation of the circumferential trenches 200 around the MEMS structures 122, in particular micromirrors, the MEMS structures 122 are connected to the underside of the MEMS chip 120 only via a silicon bridge 202.

[0066] After the release process in the wafer 100, each of the MEMS chips 120 is assigned at least one laterally extending handling surface 170, 172. This is connected to the right side of the MEMS chip 120 by means of a bridging bridge 210, while the opposite side of the MEMS chip 120 is connected to the wafer 100 via a connecting bridge 208.

[0067] Figure 2.4 shows that the composite according to Figure 2.3 is rotated and rests on a surface with the protective wafers 194, which protect the front side 192 of the wafer 100. This makes electrical contacts located on the back side 206 of the wafer 100 accessible. To perform an electrical test in the wafer composite 214, the wafer 100 can therefore rest on its front side 192 and be electrically contacted on the back side 206. Furthermore, an optical test can also be performed from the front side 192 of the wafer 100 if the at least one protective wafer 194 has a larger opening or an optical window 198 above the MEMS structures 122. In the arrangement shown in Figure 2.4, ASICs 204 can be electrically contacted on the back side 206 of the wafer 100 or on the back side of the MEMS chips 120. This is shown in the diagram in Figure 2.5.Furthermore, an electrical measurement can be carried out in the position of the wafer 100 resting on the at least one protective wafer 194. Figure 2.6 shows that a separating step is carried out by means of a laser 211, such that laser exposure over the entire area or only in certain regions leads to a separation (debonding) of the temporary bond of the at least one protective wafer 194. After this, the adhesive force of the bond is greatly reduced. Figure 2.6 also shows that in this "upside down position" of the wafer assembly 214, in particular on the right outer side, the MEMS chip 120, including its ASIC 204 accommodated on the backside metallization 190 and including the handling surfaces 170, 172 connected to it laterally via the bridge web 210, can be lifted off the at least one protective wafer 194. Previously, the connecting web 208 is separated by using the laser 211 and by a separating cut carried out by it.

[0068] In an analogous manner, the assembly shown in Figure 2.6, comprising a MEMS chip 120, an ASIC 204 arranged on the rear-side metallization 190 and a handling surface 170, 172 still laterally connected to the latter, is removed.

[0069] The illustration in Figure 2.6 further shows that the individual MEMS chips 120, together with their handling surfaces 70, 172, are successively removed from the respective protective wafer 194 and transferred to a substrate 160, and the connecting webs 208.1 to 208.8 (see also illustration in Figure 4) are separated on the front side 192 of the wafer 100. As Figure 2.6 further shows, the individual MEMS chips 120, together with the at least one laterally extending handling surface 170, 172, represented by the remaining material of the wafer 100, are held within the wafer assembly 214 by the residual holding force of the temporary bond.

[0070] According to Figures 2.7 and 2.8, the assembly separated from the at least one protective wafer 194 in Figure 2.6, comprising the MEMS chip 120 including its MEMS structures 122, the ASIC 204, and the handling surface 170, 172 connected laterally via the bridge web 210, is rotated and placed on top of the substrate 160 according to Figure 2.7. The MEMS chip 120 is removed from the wafer assembly 214 according to Figure 2.6 using a pick-and-place tool (not shown in detail here) together with the L-shaped handling surfaces 170, 172 extending laterally into the plane of the drawing and is placed on top of the substrate 160. Subsequently, as shown in Figure 2.8, the handling surface 170, 172 is separated by the action of a laser beam 211 by separating the bridge web 210 as soon as the MEMS chip 120 with the ASIC 204 accommodated on its rear side is placed on the surface of the substrate 160.After the MEMS chip 120 has been placed, the L-shaped handling surface 170, 172 is separated with minimal force by means of a separating cut performed with a laser 211. For this purpose, all bridge webs 210.1, 210.2, 210.3, 210.4, as shown in Figure 5, are separated by means of a laser 211.

[0071] Figure 2.9 shows that the individual MEMS chips 120 are removed from the wafer assembly 214 and the mounting onto the substrate 160 is repeated until several isolated MEMS chips 120 are arranged to form a MEMS array 110 on the top side of the substrate 160 while maintaining a minimized lateral distance 212 from one another.

[0072] Before the placement of the adjacent MEMS chip 120 with sensitive MEMS structures 122 formed on the top side, the separation of the laterally extending L-shaped handling surfaces 170, 172 is carried out by means of the laser 211, as shown in Figure 2.8.

[0073] Figure 3.1 shows a top view of a MEMS chip 120, on which the handling surfaces 170, 172 created during the exposure according to Figure 2.3 extend laterally. From the top view according to Figure 3.1, it can be seen that the handling surfaces 170, 172, i.e., the remaining residual material of the wafer 100, essentially form an L-shaped configuration, comprising a first leg 184 and a second leg 186 extending at right angles thereto. In Figure 3.1, the handling surfaces 170, 172 are connected to the first chip side 178 and to the second chip side 180 via the bridge webs 210.1, 210.2, 210.3, 210.4.

[0074] The illustration in Figure 3.2 shows that after the MEMS chips 120 have been placed on the top side of the substrate 160 using the pick-and-place tool, the bridge webs 210.1, 210.2, 210.3, 210.4 are separated by applying a laser 211 or a separating cut performed thereby. Thus, according to Figure 3.2, the handling surfaces 170, 172 are separated from the MEMS chip 120 to be placed and mounted along a separation line 134.

[0075] Figure 4 shows a top view of the arrangement of MEMS chips 120 in the wafer assembly 214 as well as the connecting webs 208, with which, for example, a MEMS chip 120 located in the center of the wafer assembly 214 is connected to the adjacent MEMS chips 120. The wafer assembly 214 according to the top view in Figure 4 is held together by the connecting webs 208.1 to 208.8 formed from the material of the pre-structured silicon wafer, and also by the protective wafers 194 (not shown here). The wafer assembly 214, which comprises the respective protective wafers 194, enables the forces required for the separation step to be absorbed.

[0076] The solution proposed by the invention now makes it possible to handle MEMS chips 120 that have full-surface MEMS structures 122 on their upper side. Since these are eliminated as handling surfaces to avoid damage, the method proposed by the invention can enable handling of such a MEMS chip having full-surface MEMS structures by means of the handling surfaces 170, 172 with L-shaped edge elements. The method proposed by the invention further makes it possible to design the release step such that the MEMS chips 120, except for the connecting webs 208, 208.1 - 208.8, are also released circumferentially from the wafer composite 214 by the release process through the formation of preferably circumferential trenches 200. Even after the separation of said connecting webs 208, 208.1 - 208.8.8, the MEMS chips 120 remain within the wafer assembly 214, which is formed by the respective protective wafers 194.

[0077] Figure 5 shows that after cutting through the connecting webs 208.1 to 208.8, the MEMS chip 120 with its lateral handling surfaces 170, 172 is removed from the wafer assembly 214 and is placed on the top side of the substrate 160 by means of a pick-and-place tool not shown in detail here. Figure 5 shows that the individual MEMS chips 120, which are arranged on the upper side of the substrate 160, are arranged with a minimized distance 212 in relation to one another, which results in a very high optical fill factor in relation to the area of ​​the substrate 160. For handling the MEMS chip 120 to be placed as shown in Figure 5, the handling surfaces 170, 172 are used, which laterally enclose the MEMS chip 120 to be placed in an L-shape and are connected to it via the still existing bridge webs 210.1, 210.2, 210.3, 210.4.After the MEMS chip 120 has been placed and the minimized distance 212 has been set between the MEMS chip 120 to be placed and the MEMS chips 120 already arranged on the surface of the substrate 160, the bridge webs 210.1, 210.2, 210.3, 210.4 are separated by a laser cut 211, as shown in Figures 3.2 and 2.8, so that after repeated removal of MEMS chips 120 from the wafer assembly 214 and their mounting on the substrate 160, a MEMS array 110 is created.

[0078] Figure 6 shows an enlarged view of one of the bridge webs 210 between the MEMS chip 120 and the lateral handling surface 170, 172.

[0079] The number and position of both the connecting webs 208 within the wafer assembly 214 according to Figure 4 and the bridging webs 210 between a chip edge 234 of the MEMS chip 120, on the one hand, and an inner side of the handling surfaces 170, 172, on the other hand, can be adapted in terms of their number, position, and geometry depending on the required load during electrical testing during assembly of the ASIC 204 and during assembly of the MEMS chip 120 onto the substrate 160. The mechanical strength of both the connecting webs 208 and the bridging webs 210 is essentially determined by their thickness 216, their width 218, and their length 220. The larger these dimensions are, the more resilient the connection is, but the more difficult it is to separate the connection with a laser 211. For separating the connection, for example for cutting through the bridge webs 210.1, 210.2, 210.3, 210.4 The following methods can be used to create a crack between the chip edge 234 of the MEMS chip 120 and the inner side of the handling surfaces 170, 172: A local defect is introduced in the region of a narrowest point by creating a mechanical stress gradient using local laser energy coupling and adjacent local cooling. Consequently, the defect propagates as a crack in a mechanical stress field generated in this way and can be guided in a defined manner by the stress field.

[0080] Furthermore, there is the possibility of laser ablation of the material, ie of the silicon in the region of a narrowest point of the connecting webs 208 or the bridge webs 210 or also of melting the silicon material in the region of a narrowest point of the connecting webs 208 or the bridge webs 210.

[0081] From Figures 6 and 6.1, it can be seen that the illustrated bridge web 210 is essentially formed by a first triangle 226 and a second triangle 228. The two triangles 226, 228 overlap each other within an overlap zone 230. A base side of the first triangle 226 is directed toward the MEMS chip 120, while the base side of the second triangle 228 is directed toward the handling surface 170, 172. A height 222 of the base 224 as well as angles α and β of the two triangles 226, 228 are dimensioned such that the tips overlap and both triangles 226, 228 form a narrowest point of the bridge web 210. At the narrowest point of the overlap 230, the width 218 is present, which lies approximately on the line of an outer chip edge of the MEMS chip 120. The thickness of the bridge web 210 is designated by reference numeral 216 (see Figure 6.1).Recesses 232 may be present above and optionally also below the connection of the bridge web 210 to the MEMS chip 120 and also optionally on the connected handling surfaces 170, 172.

[0082] The geometry of the bridge web 210 shown in Figures 6 and 6.1 can be further varied via the angles α and β. The angles α and β can be used to positively influence both the mechanical stability of the bridge web 210 and, after melting, the formation of a new, energetically more favorable surface of the two separate triangle vertices.

[0083] Figures 7 and 7.1 show a top view of a bridge web 210 and its side view. The connecting webs 208 and the bridge webs 210 are separated in the region of their narrowest point by energy input by means of the laser 211. The material, which is in particular silicon material, is heated as locally as possible at this point by a laser beam 211 to such an extent that it melts and the two triangles 226, 228 are separated at the tips. A resulting rounded tip of the second triangle 228 on the MEMS chip 120 preferably lies within the outer contour of the MEMS chip 120, which is formed by the chip edge 234. It is desirable that this tip does not protrude beyond the outermost chip edge 234 so that an adjacent MEMS chip 120 can be mounted on the substrate 160 at a minimized distance 212.In order to achieve local melting, it is advantageous to limit the dimensions and / or isolate the temperature field generated by the laser beam absorption. Furthermore, it is advantageous to achieve the highest possible local absorption. Both factors lead to locally high temperatures and thus to melting due to the energy input by the laser 211. By introducing structures, in particular structuring and weakening zones 236 in the form of depressions, recesses, or holes on the surface or through the entire material thickness, the surface can be modified such that the radiation applied by the laser 211 can be better absorbed and a greater volume shrinkage occurs.Furthermore, weakened zones or structuring zones 236 simultaneously offer a reduction in thermal conductivity (thermal insulation) with respect to the MEMS chip 120 or with respect to the handling surfaces 170, 172, respectively, so that a further local increase in temperature can be achieved. In Figure 7.1, for example, the bridge web 210 (210.1, 210.2, 210.3, 210.4) shown there has a cross-sectional reduction 238. Such a cross-sectional reduction 238 represents a reduction in the material density. During melting by application of the laser 211, the material is redistributed in such a way that the surface area is minimized. This causes the material to shrink in such a way that newly formed outer contours of the connecting webs 208 or the bridge webs 210 are rounded off in the region of the triangle tips of the triangles 226, 228 and each retreat towards the base, as is the case, for example, in Figures 8 and 8.1 is shown.

[0084] Figures 8 and 8.1 show that in plan view the bridge web 210 (210.1 ,

[0085] 210.2, 210.3, 210.4) has a gap 237, meaning that the handling surface 170, 172 is no longer connected to the MEMS chip 120. Since the first and second triangles 226, 228 each have rounded, nose-shaped tips, the gap 237 is created. The remaining part of the connecting or bridging web 208, 210, in particular the rounding on the second triangle 228, which is assigned to the MEMS chip 120, lies within the chip edge 234, so that when mounting a MEMS chip 120 thus exposed, as shown in Figure 8, said minimized distance 212 can be realized from a MEMS chip 120 to be placed adjacently.

[0086] Figure 8.2 shows that a material reduction or a weakening zone 236 can be achieved not only by structuring, as shown in Figures 7 and 7.1, but also by a cross-sectional reduction 238 in the form of a material reduction. If energy is applied here via the application of a laser 211, the highest energy input and thus melting occurs precisely at this point, so that, as described in connection with Figure 8, the gap 237 forms between the first triangle 226 and the second triangle 228, as shown in Figure 8.

[0087] The shortening of the first and second triangles 226, 228 creates the necessary free space to remove the individual MEMS chips 120 from the wafer assembly 214 or to be able to remove the handling surfaces 170, 172 from the remaining silicon material of the wafer assembly 214 after separation.

[0088] Figure 9 shows a further embodiment in which the region of the narrowest point of the bridge web 210.1, 210.2, 210.3, 210.4 also has the smallest thickness (cf. a first thickness 244 and a second thickness 246). Such a design of the bridge webs 210.1, 210.2, 210.3, 210.4 allows for further limitation of the temperature field. Furthermore, a narrowest point can be formed by a first radius 240 and a second radius 242, as shown in connection with Figure 9. All embodiments with regard to the arrangement of the weakening and structuring zones 236 or the cross-sectional reduction 238, as well as the formation of the narrowest point between the triangles 226, 228, can be combined with one another. The illustration according to Figure 10 schematically shows the sequence of the method proposed according to the invention, starting from the provision step 300.After the provision step 300, the at least one protective wafer 194 is attached 310 to the front side 192 of the wafer 100, which is preferably provided as a pre-structured silicon wafer. During the release step 320, circumferential trenches 200 are created around the MEMS chips 120 or around the MEMS structures 122 protected against damage. After a rotation of the wafer assembly 214, a back-side processing 330 of the pre-structured silicon wafer 100 on its back side 206 and an electrical contacting step are carried out, which is followed by the assembly 340 and back-side metallization 190 with ASICs 204.Step 350 represents a removal step in which the temporary bond connections between the wafer assembly 214 and the at least one protective wafer 194 are weakened. In step 360, the wafer assembly 214 is separated by removing individual released MEMS chips 120 together with handling surfaces 170, 172 attached laterally to one or both sides 178, 180 via the bridge webs 210. In step 370, the MEMS chips 120 are placed on the top side of the substrate 160, followed by a removal 380 of the handling surfaces 170, 172 from the chip edge 234 of the MEMS chips 120. The removal of the individual MEMS chips 120 from the wafer assembly 214 is accompanied by the construction 390 of a MEMS array 110 from a plurality of MEMS chips 120 on the surface of the substrate 160, wherein minimized distances 212 can be realized between the individual MEMS chips 120, which in turn promote a high optical fill factor.

[0089] Furthermore, the invention relates to the use of the method for producing and handling MEMS chips 120 and their placement on the surface of a substrate 160, wherein minimized distances 212 are maintained between the individual MEMS chips 120 to be placed.

[0090] The invention is not limited to the embodiments described here and the aspects highlighted therein. Rather, numerous modifications are possible within the scope of the claims, which are within the scope of one skilled in the art.

Claims

Claims 1 . Method for handling MEMS chips (120) with MEMS structures (122), in particular sensitive MEMS structures (122), which are arranged on a top side (174) of the MEMS chips (120), in particular over the entire surface, with the following method steps: a) providing (300) at least one handling surface (170, 172), which is preferably formed laterally on the MEMS chips (120) and is connected to the MEMS chip (120) via connecting webs (208), b) fastening (310) at least one protective wafer (194) on a front side (192) of the wafer (100), whereby the sensitive MEMS structures (122) are protected from damage during the execution of process steps (320 to 360), c) isolating and singulating MEMS chips (120) during the processing of the wafer (100) and producing laterally on the MEMS chips (120) extending handling surfaces (170, 178) and d) force-free removal (380) of the handling surfaces (170,172) after processing and placing the MEMS chips (120) of them., 2. Method according to claim 1, characterized in that the MEMS chips (120) comprise sensitive MEMS structures (122) for microelectromechanical systems and a release (320) of the MEMS structures (122) is carried out before singulating the MEMS chips (120) in the wafer (100).

3. Method according to claims 1 and 2, characterized in that the MEMS structures (122), in particular the sensitive MEMS structures (122), MEMS structures (122) for one or more MEMS components, such as MEMS sensors and MEMS actuators, for example MEMS inertial sensors, MEMS pressure sensors, MEMS Microphones, MEMS micromirrors and / or MEMS resonators.

4. Method according to claims 1 to 3, characterized in that the MEMS chips (120) of the wafer (100) are at least partially electrically connected to one or more components (ASICs) (204) before the wafer (100) is separated (360).

5. Method according to claims 1 to 4, characterized in that according to method step a) a pre-structured silicon wafer (100) with solderable backside metallization (190) is provided as the wafer (100).

6. Method according to claims 1 to 5, characterized in that according to method step b) at least one protective wafer (194) is attached to a front side (192) of the wafer (100) by means of a temporary bond connection, which protective wafer has a cavity and / or at least one opening (196) or an optical window (198).

7. Method according to claims 1 to 6, characterized in that according to method step c) during the exposure (320) silicon structures are dissolved out by etching in SFe and / or XeF2 and outwardly exposed protective oxide is removed by HF gas phase etching.

8. Method according to claims 1 to 7, characterized in that according to method step c) during the release (320) the MEMS chips (120) in the wafer (100) are exposed laterally by a circumferential trench (200).

9. Method according to claims 1 to 8, characterized in that after carrying out method step c), the MEMS chips (120) remain locally connected to the wafer (100) via connecting webs (208, 208.1 - 208.8) and a protective layer on the solderable backside metallization (190) of the wafer (100) is removed.

10. Method according to claims 1 to 9, characterized in that the wafer (100) lying on the front side (192) protected by at least one protective wafer (194) is (206) is electrically contacted and ASICs (204) are applied to the back (206), electrically contacted and measured.

11. Method according to claims 1 to 10, characterized in that by means of a full-area or region-wise laser exposure (211) the temporary bond connections of the at least one protective wafer (194) are released and the connecting webs (208, 208.1 - 208.8) are separated.

12. Method according to claims 1 to 11, characterized in that the MEMS chips (120) separated according to method step c) are temporarily held in the wafer assembly (214) together with the handling surfaces (170, 172) and are removed from the wafer assembly (214) by means of a pick-and-place tool.

13. Method according to claims 1 to 12, characterized in that the separated MEMS chips (120) are placed (370) on the substrate (160) and a substantially force-free removal (380) of the handling surfaces (170, 172) from the MEMS chips (120) takes place.

14. Method according to claims 1 to 13, characterized in that a removal of MEMS chips (120) from the wafer assembly (214) and a mounting of the MEMS chip (120) on the substrate (160) are repeated, so that a number of MEMS chips (120) are assembled at a minimized distance (212) from one another to form a MEMS array (110).

15. MEMS chip (120) with at least one handling surface (170, 172), manufactured according to the method according to one of claims 1 to 14, characterized in that bridging webs (210, 210.1 - 210.4) between a chip edge (234) of the MEMS chip (120) and the at least one handling surface (170, 172) are formed substantially by a first triangle (226) and a second triangle (228) which have an overlap zone (230) in which either a structuring or weakening zone (236) or a cross-sectional reduction (238) of the silicon material of the wafer (100) is present.

16. Use of the method according to one of claims 1 to 14 for the production of MEMS chips (120) and for their placement (370) on a substrate (160) while maintaining minimized distances (212) from one another.

Images