A manufacturing method of a MEMS suspended sensitive structure

Abstract

The present invention provides a manufacturing method for a MEMS suspended sensitive structure, which is characterized by the following steps: (a) Select a double-polished silicon wafer as the substrate wafer, and fabricate a movable cavity and support anchors thereon; (b) Bond a SOI wafer on the substrate wafer; (c) Grow a silicon dioxide layer; (d) Fabricate a silicon dioxide etching mask layer; (e) Prepare an etching window; (f) Perform the first deep silicon etching; (g) Remove the silicon dioxide etching mask layer; (h) Perform the second deep silicon etching; (i) Remove the photoresist and the oxidation layer etching mask. The steps of the present invention are simple and convenient to implement, effectively improve or solve the heat dissipation problem in etching and release, and ensure the integrity in the deep silicon etching and release process.

Description

Technical Field:

[0001] The present invention relates to the field of semiconductor technology, and specifically to a method for fabricating a MEMS suspended sensitive structure. Background Art:

[0002] Etching of the sensitive structure is a key process step in the fabrication of MEMS devices. In the mainstream process scheme, the wafer of the sensitive structure layer and the substrate wafer with support anchors and movable cavities are bonded together by silicon-silicon bonding. After thinning the wafer of the sensitive structure layer to the required thickness, deep silicon etching technology is used to etch through the sensitive structure layer to form a suspended and movable sensitive structure.

[0003] During the etching process of the suspended sensitive structure, a large amount of heat is generated on the surface of the sensitive structure by ion bombardment and chemical reactions. The heat is transferred from the silicon sensitive structure to the anchor points and then to the silicon substrate with helium cooling. When the heat generated during etching on the structure cannot be dissipated through the anchor points in time, heat accumulation will occur on the sensitive structure, causing local overheating (localheating). Due to the Lag effect of deep silicon etching, the large etching opening of the sensitive structure is quickly etched through, and the heat conduction path of the etching gradually becomes fewer. When the smallest etching opening size is gradually etched completely, the heat conduction efficiency of the sensitive structure is the worst, and the problem of local overheating is the most significant. Excessive temperature will cause the imbalance of the "passivation-etching" effect in deep silicon etching, resulting in etching damage to the sensitive structure, and in severe cases, the device cannot be used.

[0004] The heat conduction efficiency of the suspended sensitive structure has a great relationship with the structure design and the wafer structure. For example, the longer and thinner the beam structure, the lower the heat conduction rate; the larger the area of the sensitive structure, the more heat generated by ion bombardment; the thicker the buried oxide layer in the wafer structure, the worse the heat dissipation. The problem of local overheating makes it difficult to ensure the integrity of the sensitive structure. It is a common process difficulty in the deep silicon etching of sensitive structures, which greatly restricts the design and process preparation of sensitive structures and has an important impact on the development and production of MEMS devices.

[0005] In the prior art, some documents optimize the deep silicon etching process to improve heat dissipation during etching and avoid serious etching damage caused by local overheating. One method is the segmented etching process, that is, manually pausing the etching program, waiting for a certain time for cooling, and then restarting the etching program. In this way, multiple manual pauses and restarts of the etching program are performed to help dissipate heat to achieve the purpose of improving local overheating. This method is troublesome to execute, and the improvement effect of damage is general. The etching rate will be affected by "pause-restart", and the etching steepness cannot be guaranteed. Another method is to change the "passivation-etching" cycle to a "passivation-etching-waiting for cooling" cycle in the etching program, which greatly improves the heat dissipation effect and the etching efficiency, and significantly alleviates the problem of etching overheating damage. However, this method still has a certain applicability to different sensitive structures. Summary of the Invention:

[0006] The present invention aims to improve the heat dissipation efficiency of the MEMS suspended sensitive structure during deep silicon etching and prevent etching damage caused by excessive local temperature, and proposes a manufacturing method for the MEMS suspended sensitive structure.

[0007] The present application provides the following technical solutions:

[0008] A manufacturing method for a MEMS suspended sensitive structure, characterized in that it includes the following steps:

[0009] (a) Select a double-polished silicon wafer as the substrate wafer, and then use photolithography and etching on the substrate wafer to prepare a group of movable cavities and a group of support anchors;

[0010] (b) Bond a SOI wafer on the substrate wafer, and remove the top silicon and buried oxide layer of the SOI wafer to retain the substrate layer as the sensitive structure layer;

[0011] (c) Grow a silicon dioxide layer on the surface of the sensitive structure layer by PECVD or thermal oxidation method;

[0012] (d) Use photolithography to form a pattern on the silicon dioxide layer grown in (c), and then etch the other silicon dioxide layers to prepare a silicon dioxide etching mask layer corresponding to the pattern on the sensitive structure layer, and the silicon dioxide etching mask layer corresponds to the movable cavity;

[0013] (e) Apply a photoresist layer on the sensitive structure layer, the photoresist layer covers the silicon dioxide etching mask layer, and then use photolithography technology to make a pattern on the surface of the photoresist layer. A group of etching windows are formed in the pattern, and the group of etching windows includes at least two widths of etching windows. Then, etch the photoresist layer downward according to the etching windows to expose the silicon and the silicon dioxide etching mask layer on the surface of the sensitive structure layer;

[0014] (f) Use deep silicon etching to perform a first deep silicon etching of a certain depth on the silicon of the sensitive structure layer according to the etching windows prepared in (e);

[0015] (g) Use a silicon dioxide etching machine to remove the silicon dioxide etching mask layer along the etching windows;

[0016] (h) Use deep silicon etching again to perform a second deep silicon etching of a certain depth on the silicon of the sensitive structure layer according to the etching windows prepared in (e) until it is connected to the movable cavity;

[0017] (i) Remove the photoresist and the oxide layer etching mask to complete the release of the sensitive structure.

[0018] Based on the above technical solutions, the following further technical solutions may also be available:

[0019] The set of active cavities in step (a) includes a plurality of large-area active cavities and a plurality of small-area active cavities. The small-area active cavities are located between two large-area active cavities; the support anchor points are located between the small-area active cavities and the large-area active cavities.

[0020] The set of etching windows in step (e) includes etching windows with three different widths: large, medium, and small. The large-width etching windows are located above the silicon dioxide etching mask layer. The medium-width etching windows and the small-width etching windows are located between two large-width etching windows, and a number of small-width etching windows are located between two medium-width etching windows.

[0021] Advantages of the invention:

[0022] The steps of the present invention are simple and easy to implement, effectively improving or solving the heat dissipation problem in etching and release, and ensuring the integrity in the deep silicon etching and release process.

[0023] Especially in the first deep silicon etching, due to the presence of the silicon dioxide etching mask layer, the large-width etching windows in this area are not etched. At this time, effective heat dissipation can be carried out during the etching of the medium-width and small-width etching windows. Then, in the second deep silicon etching, the large, medium, and small-width etching windows can be etched through simultaneously, or the large-width etching windows are etched through slightly later than the medium-width and small-width etching windows. In this way, the heat generated during the first and second deep silicon etching processes has a good path for heat dissipation, avoiding etching damage caused by excessive local temperature.

[0024] Effectively improving or solving the heat dissipation problem in the etching and release of MEMS sensitive structures, ensuring the integrity of the sensitive structures in the deep silicon etching and release process, especially suitable for the design of sensitive structures with poor heat dissipation, with remarkable effects, and of great significance to the design and development of MEMS devices. Description of the drawings:

[0025] Figure 1 is a schematic structural diagram after completing step (a);

[0026] Figure 2 is a schematic structural diagram after completing step (b);

[0027] Figure 3 is a schematic structural diagram after completing step (d);

[0028] Figure 4 is a schematic structural diagram after completing step (e);

[0029] Figure 5 is a schematic structural diagram after completing step (f);

[0030] Figure 6It is a schematic structural diagram after completing step (g);

[0031] Figure 7 It is a schematic structural diagram after completing step (h);

[0032] Figure 8 It is a schematic structural diagram after completing step (i). Specific implementation manner:

[0033] As Figure 1-8 shown, a manufacturing method of a MEMS suspension sensitive structure is characterized in that it includes the following steps:

[0034] (a) Select a double-polished silicon wafer as the substrate wafer 1, with a thickness of 400 µm. On the upper surface of the substrate wafer, a group of movable cavities and a group of support anchors 3 are prepared by photolithography and etching. The group of movable cavities includes two large-area movable cavities 2 and two small-area movable cavities 2a.

[0035] The two small-area movable cavities 2a are located between the two large-area movable cavities, and the two small-area movable cavities 2a are spaced apart. The silicon between the small-area movable cavity and the large-area movable cavity is the support anchor 3. The depth of the group of movable cavities is 20 µm.

[0036] (b) Bond a piece of SOI wafer on the substrate wafer, with a thickness of 300 µm., and remove the top silicon and buried oxide layer of the SOI wafer, retaining the substrate layer as the sensitive structure layer 4. The thickness of the sensitive structure layer 4 is 50 µm.

[0037] (c) On the surface of the sensitive structure layer 4, grow a silicon dioxide layer on the surface of the sensitive structure layer by thermal oxidation, with a thickness of 4000 Å.

[0038] (d) Use photolithography to form a pattern on the silicon dioxide layer grown in (c), and then etch the other silicon dioxide layers to prepare a silicon dioxide etching mask layer 5 corresponding to the pattern on the sensitive structure layer. The silicon dioxide etching mask layer 5 corresponds to the large-area movable cavity 2 and is located in the area above the large-area movable cavity 2.

[0039] (e) Apply a photoresist layer 6 on the sensitive structure layer 4. The photoresist layer 6 covers the silicon dioxide etching mask 5 layer, and then use photolithography technology to make a pattern on the surface of the photoresist layer, and a group of etching windows are formed in the pattern.

[0040] The group of etching windows includes wide-width etching windows x located above the silicon dioxide etching mask layer 5. A number of medium-width etching windows y and narrow-width etching windows z are distributed between the two wide-width etching windows x.

[0041] There are three small-width etching windows z distributed between two adjacent medium-width etching windows y on one side of one large-width etching window x. The one large-width, two medium-width, and three small-width etching windows correspond to a large-area active cavity 2. There is also another medium-width etching window y correspondingly distributed in the small-area active cavity 2a region on one side of the large-area active cavity 2.

[0042] Etch the photoresist layer downward according to the etching window, so as to expose the silicon and silicon dioxide etching mask layer 5 on the surface of the sensitive structure layer 4.

[0043] (f)Use deep silicon etching to perform the first deep silicon etching of a certain depth on the silicon of the sensitive structure layer 4 according to the etching window prepared in (e). During this deep silicon etching, due to the barrier of the silicon dioxide etching mask layer 5, the large-width etching window x cannot be etched downward. The etching depth of the small-width etching window z in this deep silicon etching is 26 µm, and this depth is confirmed according to the etching rate experiment.

[0044] (g)Use a silicon dioxide etching machine to remove the silicon dioxide etching mask layer 5 downward along the large-width etching window x.

[0045] (h)Use deep silicon etching again to perform the second deep silicon etching of a certain depth on the silicon of the sensitive structure layer 4 according to the etching window prepared in (e) until it is connected to the active cavity. Since the silicon dioxide etching mask layer 5 has been removed during this deep silicon etching, the large-width etching window x will also be etched downward. And because the etching speed of the large-width etching window x is greater than that of the medium-width and small-width etching windows y and z, the three-width etching windows can be etched through simultaneously or the large-width etching window can be etched through slightly later. Thus, it effectively ensures that the heat generated during the first and second deep silicon etching processes has a good path for heat dissipation, avoiding etching damage caused by excessive local temperature.

[0046] (i)Remove the photoresist and the oxide layer etching mask to complete the release of the sensitive structure.

Claims

1. A manufacturing method of a MEMS suspension sensitive structure, characterized in that: It includes the following steps: (a) Select a double-polished silicon wafer as the substrate wafer, and then use photolithography and etching on the substrate wafer to fabricate a set of movable cavities and a set of support anchors; (b) Bond a piece of SOI wafer on the substrate wafer, and remove the top silicon and buried oxide layer of the SOI wafer, retaining the substrate layer as the sensitive structure layer; (c) Grow a silicon dioxide layer on the surface of the sensitive structure layer by PECVD or thermal oxidation; (d) Use photolithography to pattern the silicon dioxide layer grown in (c), and then etch the other silicon dioxide layers to fabricate a silicon dioxide etching mask layer corresponding to the pattern on the sensitive structure layer. The silicon dioxide etching mask layer corresponds to the movable cavities; (e) Apply a photoresist layer on the sensitive structure layer. The photoresist layer covers the silicon dioxide etching mask layer, and then use photolithography technology to pattern the surface of the photoresist layer. A set of etching windows are formed in the patterning. The set of etching windows includes at least two types of etching windows with different widths. Then, etch the photoresist layer downward according to the etching windows to expose the silicon and the silicon dioxide etching mask layer on the surface of the sensitive structure layer; (f) Use deep silicon etching to perform a first deep silicon etching of a certain depth on the silicon of the sensitive structure layer according to the etching windows obtained in (e); (g) Use a silicon dioxide etching machine to remove the silicon dioxide etching mask layer along the etching windows; (h) Use deep silicon etching again to perform a second deep silicon etching of a certain depth on the silicon of the sensitive structure layer according to the etching windows obtained in (e) until it communicates with the movable cavities; (i) Remove the photoresist and the oxide etching mask to complete the release of the sensitive structure; The set of etching windows in step (e) includes three different widths of etching windows (x, y, z) of large, medium, and small sizes. The large-width etching window (x) is located above the silicon dioxide etching mask layer. The medium-width etching window (y) and the small-width etching window (z) are located between two large-width etching windows (x), and several small-width etching windows (z) are located between two medium-width etching windows (y).

2. The manufacturing method of a MEMS suspension sensitive structure according to claim 1, wherein: The set of movable cavities in step (a) includes multiple large-area movable cavities and multiple small-area movable cavities. The small-area movable cavities are located between two large-area movable cavities; the support anchors are located between the small-area movable cavities and the large-area movable cavities.

Images