Manufacturing method of microstructure

Abstract

A method for producing a microstructure comprising silicon nitride (Si3N4) is provided. The method includes the steps of etching a sacrificial layer of silicon dioxide (SiO2) using hydrogen fluoride (HF) vapor and then removing a residual layer formed when the HF vapor also etches the silicon nitride (Si3N4). The residual layer comprises silicon, and various techniques are disclosed for removing such a layer. The techniques can be applied simultaneously or sequentially to the microstructure. Thus, the described method produces a microstructure with reduced levels of silicon residue compared to techniques known in the art.

Description

[Technical field]

[0001] The present invention relates to a method for use in the fabrication of microstructures, typically in the form of microelectromechanical systems (MEMS) that require the removal of material from a substrate or other deposited material. In particular, the present invention relates to an improved method for fabricating microstructures that uses a step of etching a silicon dioxide sacrificial layer with hydrogen fluoride (HF) vapor. [Background technology]

[0002] Many materials are known to be used during the fabrication of microstructures; silicon, silicon dioxide (SiO2), silicon nitride (Si3N4), aluminum, photoresist, to name just a few. Some of these materials are used as sacrificial materials, while others are used to define and form the microstructures. It is not uncommon for multiple sacrificial etching steps to be employed during the fabrication of microstructures. For example, a film is used as a mask during an initial sacrificial etching process and then etched as a subsequent sacrificial layer. For this reason, it is highly desirable to have high etch selectivity between the sacrificial layer and the surrounding material in any removal etch.

[0003] During the fabrication of a microelectromechanical structure (MEMS) 1 of the type shown in Figure 1, a sacrificial layer 2 and a structure-defining layer 3 are first deposited on a substrate 4. An etching process is then used to remove the sacrificial (i.e. unwanted) regions or layers 2.

[0004] One of the most common materials used as the sacrificial layer 2 is silicon dioxide, which is etched using hydrogen fluoride (HF) vapour (see, for example, GB Patent No. 2,487,716). HF vapour etching is a plasma-free chemical etch and is described by the following reaction: TIFF2024522062000002.tif30170

[0005] As shown in equation (1), water (H2O) ionizes HF vapor, and the ionized HF vapor (HF2 - ) etches silicon dioxide (SiO2) using water (H2O) as a catalyst. It is clear from equation (2) that water (H2O) is also produced from the etching reaction itself.

[0006] Hydrogen fluoride (HF) vapor etching of silicon dioxide 2 is known to be highly selective to many common structure-defining layers 3. For example, the theoretical selectivity to silicon and aluminum is high, and no etching or corrosion is expected. Silicon nitride (Si3N4) is another material often employed as a mask or structure-defining layer 3 in the fabrication process of MEMS1, although its selectivity to silicon dioxide 2 is not as high as that of silicon or aluminum.

[0007] As well as being used in the manufacture of MEMS 1, silicon dioxide 2 and silicon nitride 3 layers are also present in semiconductor devices. As such, it is also known to use hydrogen fluoride (HF) vapor etching techniques to form air gap structures in multi-level metal structures used in standard semiconductor devices (see U.S. Pat. No. 7,211,496).

[0008] It is widely accepted that for HF vapor etching to proceed at a usable etch rate, for example faster than 30 nm / min, a condensed fluid layer 5 must be present on the surface to be etched (see, for example, Journal of Vacuum Science and Technology A, 10(4) July / Aug 1992 entitled "Mechanisms of the HF / H2O vapor phase etching of SiO2" in the name of Helms et al.). Of all the compounds involved in the HF vapor phase etching process mentioned above, water (H2O) has the lowest vapor pressure and therefore forms the basis of the condensed fluid layer 5. EP 2046677 discloses how controlling the formation and composition of the condensed fluid layer 5 is important in managing the HF vapor etching of silicon dioxide. Precise etch control is achieved by performing the HF etch in a vacuum chamber and controlling the chamber pressure, temperature, and gas flow into the chamber. Other parameters that affect HF vapor etching are the composition of the silicon dioxide layer to be etched and its deposition method.

[0009] Chemical vapor deposition (CVD) processes are typically employed to deposit silicon dioxide (SiO2) layers 2 and silicon nitride (Si3N4) layers 3 on a substrate 4. In these processes, chemical precursors (a silicon source on the one hand and an oxygen or nitrogen source on the other hand) react to deposit a silicon dioxide layer 2 or silicon nitride layer 3 on the substrate 4. The most common of these processes is plasma-enhanced CVD (PECVD), as this process allows deposition to occur at low temperatures below 450°C.

[0010] It is known in the art to deposit a silicon nitride layer 3 to make such a layer silicon-rich in order to increase its selectivity with the sacrificial silicon dioxide layer 2. Even with this increased selectivity, it has been found that the HF vapor etching process etches the silicon nitride layer 3, albeit at a much slower rate than the silicon dioxide layer 2. It is known that the HF etching of the silicon nitride layer 3 can leave a solid residual layer on the MEMS 1, which contains ammonium salts. The solid residues thus formed are obviously highly undesirable in the manufacture of MEMS or semiconductor microstructures. Summary of the Invention

[0011] It is therefore an object of one embodiment of the present invention to provide a method for producing microstructures comprising silicon nitride that employs HF etching of a sacrificial silicon dioxide layer and reduces the level of solid residual layers or particles compared to the above techniques known in the art.

[0012] According to a first aspect of the present invention, there is provided a method of manufacturing a microstructure comprising silicon nitride (Si3N4), the method comprising: - selectively etching a sacrificial layer of silicon dioxide (SiO2) using hydrogen fluoride (HF) vapor; - after the etching with HF vapor is completed, removing the solid silicon residual layer formed on the silicon nitride (Si3N4) by etching the silicon nitride (Si3N4) with HF vapor.

[0013] The steam etching of the sacrificial silicon dioxide (SiO2) layer and the removal of the solid silicon residual layer can be performed sequentially in separate processing chambers.

[0014] Optionally, removing the residual layer comprises reacting the solid silicon residual layer with a first additional gas.

[0015] Optionally, removing the residual layer includes reacting the solid silicon with hydrogen gas or a hydrogen compound to produce silane (SiH4).

[0016] Alternatively, removing the residual layer can include reacting solid silicon with oxygen gas or an oxygen compound to produce silicon dioxide (SiO2). Additionally, removing the residual layer can include etching the silicon dioxide (SiO2) using hydrogen fluoride (HF) vapor.

[0017] Further alternatively, removing the residual layer includes reacting solid silicon with fluorine gas or a fluorine compound to produce silicon tetrafluoride (SiF4).

[0018] Further alternatively, removing the residual layer may include etching the solid silicon with xenon difluoride (XeF2) vapor.

[0019] The method of manufacturing a microstructure may further include removing by-products formed during removal of the solid silicon residual layer using a vacuum pump system.

[0020] Most preferably, the microstructure comprises a microelectromechanical system (MEMS). Alternatively, the microstructure comprises a semiconductor device. [Brief description of the drawings]

[0021] Various embodiments of the present invention will now be described, by way of example only, with reference to the drawings in which: [Figure 1] FIG. 1 shows a schematic diagram of HF vapor etching of a MEMS that includes a silicon dioxide layer located between a substrate and a structure-defining layer. [Diagram 2] FIG. 2 shows a schematic diagram of a process chamber system suitable for performing HF vapor etching of the MEMS of FIG. [Diagram 3] FIG. 3 shows a schematic diagram of the MEMS of FIG. 1 after an HF vapor etching process. [Figure 4] FIG. 4 shows a flow chart of a method for manufacturing a MEMS according to the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0022] Figure 2 shows a schematic diagram of an etcher 6 suitable for etching the MEMS 1 of Figure 1. The etcher 6 can be seen to include an etching chamber 7 to which six input lines 8, 9, 10, 11, 12, 13 and an output vacuum line 14 are connected.

[0023] Within the etching chamber 7 is a temperature controlled pedestal 15 suitable for placing the MEMS structure 1 to be etched within the etching chamber 7. Fluids supplied from six input lines 8, 9, 10, 11, 12, 13 enter the interior volume of the etching chamber 7 via a fluid injection system 16 located within a lid 17 of the etching chamber 7.

[0024] The pedestal 15 on which the MEMS 1 is disposed is controlled by a temperature controller to a pedestal temperature T P This temperature may be set and maintained at a temperature above or below room temperature, with the specific temperature being selected to optimize the etching process (typically between 5 and 25° C.). Additionally, the walls of the etching chamber 7 are heated during the etching process, typically to about 20 to 70° C.

[0025] The pressure of the etching gas in the etching chamber, P C is monitored by a chamber pressure controller 18. The pressure controller 18 also incorporates gas flow controllers which are employed to provide a means of controlling the pressure within the etch chamber 7 by controlling the operation of a vacuum pumping system 19 located on the output vacuum line 14.

[0026] HF vapor 20 is controllably supplied to the etching chamber 7 by a first input line 8 via a regulator 21 and a first mass flow controller (MFC) 22 .

[0027] A controlled amount of water is supplied to the etching chamber 7 via a second input line 9. In particular, a liquid fluid controller (LFC) 23 and a vaporizer 24 disposed within the second input line 9 are employed to generate a controlled level of water vapor from a water reservoir 25. The flow of nitrogen from a nitrogen gas source 26 to the vaporizer 24 is controlled by a second MFC 22. A nitrogen carrier gas is used to transport the water vapor into the interior volume of the etching chamber 7 via the fluid injection system 16.

[0028] A third input line 10, a fourth input line 11 and a fifth input line 12 provide means for connecting additional gas sources 27, 28, 29, such as hydrogen (H2) or hydrogen compounds, oxygen (O2) or oxygen compounds, or fluorine (F2) or fluorine compounds, to the internal volume of the etching chamber 7. Control of those gas flows is likewise provided by mass flow controllers (MFCs) 22.

[0029] Xenon difluoride (XeF2) vapor 30 is controllably supplied to the etching chamber 7 by a sixth input line 13 via a second regulator 21 and a mass flow controller (MFC) 22. The flow of XeF2 vapor 30 may also be controlled by a carrier gas flowing through a bubbler or vessel in which the XeF2 is contained.

[0030] A computer controller 31 is used to automate the regulation of the various components and parameters of the etch chamber 7, such as the supply of nitrogen carrier gas, HF vapor, chamber temperature and pressure, etc.

[0031] In order to proceed with the described etching method, it is necessary to have diagnostics that allow accurate monitoring of the condensed fluid layer 5. As mentioned above, the physical properties of the condensed fluid layer 5 directly affect the etch rate of the MEMS 1, and therefore monitoring the etch rate provides a direct diagnostic of the physical properties of the condensed fluid layer 5.

[0032] In practice, the etch rate can be monitored in a variety of ways, for example, by monitoring the levels of by-products produced, by directly monitoring the wafer etching, or by monitoring changes in chamber conditions.

[0033] Figure 3 shows a schematic diagram of the MEMS 1 of Figure 1 after an HF vapor etch process performed in etcher 6 of Figure 2. As shown, once the HF process is complete, the silicon dioxide layer 2 is removed, thereby freeing the silicon nitride (Si3N4) layer 3 from the substrate 4 so that it can be manipulated as desired. Unfortunately, a solid residual layer, generally designated 32, is present on the exposed surface of the MEMS 1. The solid residual layer 32 has the appearance of a random distribution of particulate residue.

[0034] Of the chemical precursors employed in the aforementioned PECVD process to deposit the silicon nitride layer 3, the most commonly used silicon source is silane (SiH4).

[0035] The applicant has found that the residual layer 32 produced during HF vapor etching of the MEMS 1 surprisingly often contains silicon impurities, instead of the expected ammonium salts. This silicon impurity is found to be present regardless of the silicon source employed in the deposition process of the silicon nitride (Si3N4) layer 3. However, the presence of silicon impurities is found to be increased in the residual layer 32 when PECVD conditions are used to produce a silicon-rich silicon nitride (Si3N4) layer 3, highlighting the source of the solid residual layer 32 as the HF vapor etching of the silicon nitride (Si3N4) layer 3. In such a situation, the HF vapor etches the silicon nitride (Si3N4) layer 3, but not the silicon contaminants contained therein, so that the silicon contaminants remain as solid residues. Until now, the silicon-based solid residual layer 32 has been completely ignored in the art, since the presence of silicon in post-etch analysis of the MEMS 1 has always been attributed by those skilled in the art to silicon present in other areas of the device (e.g., substrate 4).

[0036] Various methods for manufacturing microstructures such as semiconductor devices or MEMS 1, including the removal of the solid silicon residual layer 32 according to the present invention, will now be described with reference to FIG.

[0037] The process includes the step of performing an HF vapor etch of the sacrificial silicon dioxide layer 2 as detailed with respect to Figures 1-3. The process then includes the step of removing the solid silicon residual layer 32 from the microstructure by employing one or more of the following techniques: As mentioned above, the solid silicon residual layer 32 is formed by the HF vapor also etching the silicon nitride (Si3N4) layer 3, since the selectivity with the silicon dioxide (SiO2) sacrificial layer 2 is never absolute.

[0038] Because the residual layer 32 contains silicon, a third input line 10 can be used to connect a hydrogen or hydrogen compound gas source 27 to the etching chamber 7. The hydrogen or hydrogen compound gas may be ionized, for example by a remote plasma system, before being supplied to the etching chamber 7. Alternatively, the hydrogen or hydrogen compound gas can be ionized inside the etching chamber 7 itself. The silicon in the residual layer 32 then reacts with the hydrogen to produce silane (SiH4). Because silane (SiH4) is a volatile material, it can be easily evacuated from the etching chamber 7 by the vacuum pumping system 19.

[0039] Alternatively, a fourth input line 11 can be used to connect an oxygen or oxygen compound gas source 28 to the etching chamber 7. The oxygen or oxygen compound gas may be ionized, for example by a remote plasma system, before being supplied to the etching chamber 7. Alternatively, the oxygen or oxygen compound gas may be ionized inside the etching chamber 7 itself. The silicon in the residual layer 32 then reacts with the oxygen to produce silicon dioxide (SiO2). The HF vapor etching process described above can then be repeated to remove the residual layer 32 in a similar manner as described for the sacrificial silicon dioxide layer 2. The by-products of this second HF vapor etching process can likewise be simply evacuated from the etching chamber 7 by the vacuum pumping system 19.

[0040] In yet another alternative embodiment, a fifth input line 12 can be used to connect a fluorine or fluorine compound gas source 29 to the etching chamber 7. The fluorine or fluorine compound gas may be ionized, for example by a remote plasma system, before being supplied to the etching chamber 7. Alternatively, the fluorine or fluorine compound gas may be ionized within the etching chamber 7 itself. The silicon in the residual layer 32 then reacts with the fluorine to produce silicon tetrafluoride (SiF4). Since silicon tetrafluoride (SiF4) is a volatile material, it can be easily pumped out of the etching chamber 7 by the vacuum pump system 19.

[0041] A fourth technique that can be employed to remove the solid silicon residual layer 32 is to perform a xenon difluoride (XeF2) vapor etching process in the etching chamber 7. Here, a sixth input line 13 is used to connect xenon difluoride (XeF2) vapor 30 to the etching chamber 7. Both commonly owned European Patents 1,766,665 and 2,480,493 disclose techniques for etching silicon with xenon difluoride (XeF2) vapor that can be performed by adapting the etching apparatus 6 of FIG.

[0042] It will be appreciated that the above-mentioned methods for removing the solid silicon residual layer 32 can be problematic because there are usually other exposed areas of silicon material that form the MEMS 1 or semiconductor device to operate. That is, any of the above-mentioned techniques for removing the silicon residual layer 32 are expected to also remove silicon in their surrounding areas. However, applicants have discovered that the etch rate of the silicon residual layer 32 is much higher than the etch rate of silicon typically found in the peripheral areas of the device. Therefore, it is possible to carry out the above-mentioned techniques before the surrounding silicon areas are significantly etched. Applicants believe that the reason for this significant difference in silicon etch rate is due to the fact that the silicon in the exposed solid residual layer 32 is not a well-structured solid but rather comprises a highly amorphous porous structure.

[0043] In the event that the residual layer 32 contains anticipated ammonium salts, another technique must be employed. Ammonium salts are known to decompose at temperatures above 160° C., and applicants have realised that the ammonium salts can be evaporated by heating the MEMS 1 using a heating element within the etching chamber 7, which can then be evaporate from the etching chamber 7 by the vacuum pumping system 19.

[0044] The above-described method for forming MEMS containing silicon nitride (Si3N4) has significant advantages over systems known in the art that employ an HF vapor etching step to remove a silicon dioxide (SiO2) sacrificial layer in that it provides a means for reducing or eliminating the solid silicon residual layer that forms as a by-product of HF vapor that also etches silicon nitride (Si3N4). Most importantly, the above-described method provides a heretofore unrecognized means for removing the silicon-based residual layer.

[0045] Although the above techniques are described with particular reference to MEMS, the techniques also apply to other microstructures (e.g., semiconductor devices) including silicon nitride (Si3N4), the fabrication of which employs an HF vapor etching step to remove a silicon dioxide (SiO2) sacrificial layer.

[0046] While the above-mentioned HF vapor etching and removal of the silicon residual layer have been described as occurring in a common process chamber, those skilled in the art will appreciate that in alternative embodiments, the microstructure is transferred from a first process chamber, in which a HF vapor etch is previously performed, to a second process chamber in order to perform one or more of the above-mentioned techniques for removing the silicon residual layer.

[0047] A method for producing a microstructure comprising silicon nitride (Si3N4) is provided, the method comprising the steps of etching a sacrificial layer of silicon dioxide (SiO2) using hydrogen fluoride (HF) vapor and then removing a residual layer formed when the HF vapor also etches the silicon nitride (Si3N4). The residual layer comprises silicon, and various techniques are disclosed for removing such a layer. The techniques can be applied simultaneously or sequentially to the microstructure. Thus, the described method produces a microstructure with reduced levels of silicon residue compared to techniques known in the art.

[0048] The above description of the invention has been presented for purposes of illustration and description and is not intended to be exhaustive or to limit the invention to the precise form disclosed. The described embodiments have been chosen and described in order to best explain the principles of the invention and its practical application, thereby enabling others skilled in the art to best utilize the invention in its various embodiments with various modifications suited to the particular uses envisaged. Accordingly, further changes or modifications may be incorporated without departing from the scope of the invention as defined by the appended claims.

Claims

【Claim 1】 A method for manufacturing a microstructure containing silicon-rich silicon nitride (Si ３ N ４ ), comprising: - Etching a sacrificial layer of silicon dioxide (SiO ２ ) using hydrogen fluoride (HF) vapor, wherein the HF vapor also etches silicon-rich silicon nitride (Si₃N₄); - After the etching with the HF vapor is completed, removing the solid silicon residue formed by the etching with the HF vapor A method characterized by comprising the above steps. Claim 2 A method for manufacturing a microstructure including silicon-rich silicon nitride (Si ３ N ４ ), characterized in that vapor etching of a sacrificial layer of silicon dioxide (SiO ２ ) and removal of solid silicon residues are sequentially performed in separate processing chambers. Claim 3 A method for manufacturing a microstructure including silicon-rich silicon nitride (Si ３ N ４ ), characterized in that removing the solid silicon residue includes reacting the solid silicon residue with a first additional gas. Claim 4 A method for manufacturing a microstructure comprising silicon-rich silicon nitride (Si ３ N ４ ), wherein removing the solid silicon residue comprises reacting the solid silicon residue with hydrogen or a hydrogen compound gas to produce silane (SiH ４ ). Claim 5 A method for manufacturing a microstructure including silicon-rich silicon nitride (Si ３ N ４ ), wherein removing the solid silicon residue includes reacting the solid silicon residue with oxygen or an oxygen compound gas to produce silicon dioxide (SiO ２ ). The method is characterized by this. Claim 6 A method for manufacturing a microstructure including silicon-rich silicon nitride (Si ３ N ４ ), characterized in that removing the solid silicon residue further comprises etching silicon dioxide (SiO ２ ) using hydrogen fluoride (HF) vapor. Claim 7 A method for manufacturing a microstructure including silicon-rich silicon nitride (Si ３ N ４ ), characterized in that removing the solid silicon residue includes reacting the solid silicon residue with a fluorine or fluorine compound gas to generate silicon tetrafluoride (SiF ４ ). Claim 8 A method for manufacturing a microstructure including silicon-rich silicon nitride (Si ３ N ４ ), wherein removing the solid silicon residue comprises etching the solid silicon residue with xenon difluoride (XeF ２ ) vapor. The method is characterized by this. Claim 9 A method for manufacturing a microstructure comprising silicon-rich silicon nitride (Si ３ N ４ ), further comprising the step of removing by-products formed when removing the solid silicon residue using a vacuum pump system. Claim 10 A method for manufacturing a microstructure including silicon-rich silicon nitride (Si ３ N ４ ), characterized in that the microstructure includes a microelectromechanical system (MEMS). Claim 11 A method for manufacturing a microstructure containing silicon-rich silicon nitride (Si ３ N ４ ), characterized in that the microstructure includes a semiconductor device.

Images