Method for manufacturing a semiconductor device, and plasma processing apparatus.
By forming a protective film on the laminate surface using plasma-activated gases and controlled etching, the method addresses the selectivity challenge in GAA structure manufacturing, ensuring accurate and reliable nanowire channel formation.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- SHIBAURA MECHATRONICS CORP
- Filing Date
- 2022-09-30
- Publication Date
- 2026-06-22
AI Technical Summary
Existing methods for manufacturing semiconductor devices with a gate all around (GAA) structure face challenges in achieving the desired selectivity ratio of silicon germanium to silicon during isotropic etching, leading to reduced dimensional accuracy and reliability of nanowire channels.
A method involving the formation of a protective film on the laminate surface using radicals generated by plasma-activated gases, followed by isotropic etching with a different gas mixture, allowing controlled etching of silicon germanium layers while preserving the silicon layers, and repeated protective film formation to maintain selectivity and accuracy.
Enhances the selectivity ratio of silicon germanium to silicon, improving the dimensional and shape accuracy of nanowire channels, thereby enhancing the functionality and reliability of the semiconductor device.
Smart Images

Figure 0007877151000001 
Figure 0007877151000002 
Figure 0007877151000003
Abstract
Description
Technical Field
[0001] Embodiments of the present invention relate to a method for manufacturing a semiconductor device and a plasma processing apparatus.
Background Art
[0002] In the process of manufacturing a semiconductor device, layers with different constituent elements may be stacked, and a layer containing one element may be isotropically etched.
[0003] For example, when miniaturizing a MOSFET (Metal-Oxide-Semiconductor Field Effect Transistor), the on-off ratio becomes small. Therefore, a MOSFET having a gate all around (GAA) structure has been proposed so as to obtain a desired on-current.
[0004] In the process of manufacturing a MOSFET having a GAA structure, a plurality of layers containing silicon germanium (SiGe) and layers containing silicon (Si) are alternately stacked on a substrate, and these are sequentially etched to form a laminate in which layers containing silicon germanium and layers containing silicon are alternately stacked. Then, the layers containing silicon germanium included in the laminate are removed by isotropic etching. The layers containing silicon included in the laminate become so-called nanowire channels.
[0005] Here, when performing isotropic etching, if the selectivity of silicon germanium with respect to silicon is not increased, the amount of erosion of the layers containing silicon (nanowire channels) included in the laminate will increase. That is, when the selectivity of silicon germanium with respect to silicon is increased, only the layer to be etched (the layer containing silicon germanium) will be etched. As a result, it becomes possible to form the dimensions aimed at the layers containing silicon (nanowire channels). If it is not possible to form the dimensions aimed at the layers containing silicon (nanowire channels), the functions and reliability of the MOSFET having a GAA structure may decrease.
[0006] Therefore, there has been a need to develop a technique that can increase the selectivity ratio of one element to the other when layering layers with different elemental compositions and isotropically etching the layer containing one element. [Prior art documents] [Patent Documents]
[0007] [Patent Document 1] Japanese Patent Publication No. 2011-29503 [Overview of the project] [Problems that the invention aims to solve]
[0008] The problem that the present invention aims to solve is to provide a method for manufacturing a semiconductor device and a plasma processing apparatus that can increase the selectivity ratio of one element to the other element when stacking layers with different elemental compositions and isotropically etching a layer containing one element. [Means for solving the problem]
[0009] A method for manufacturing a semiconductor device according to an embodiment includes the steps of forming a protective film on a surface including the side surface of a laminate in which a first layer and a second layer having different elemental composition from the first layer are stacked, and isotropically etching the laminate on which the protective film is formed. The step of forming the protective film is performed using a first radical generated by exciting and activating a first gas with plasma. The step of isotropically etching the laminate is performed using a second radical generated by exciting and activating a second gas different from the first gas with plasma. In the process of isotropically etching the laminate, if it is determined that the protective film formed on the surface of the second layer has been removed, the isotropic etching is stopped and the process of forming the protective film is performed. The step of forming the protective film and the step of isotropically etching the laminate are performed in the same atmosphere. [Effects of the Invention]
[0010] According to embodiments of the present invention, a method for manufacturing a semiconductor device and a plasma processing apparatus are provided that can increase the selectivity ratio of one element to the other element when layers with different elemental compositions are stacked and a layer containing one element is isotropically etched. [Brief explanation of the drawing]
[0011] [Figure 1] This is a schematic plan view of a semiconductor device. [Figure 2] Figure 1 is a schematic cross-sectional view of the semiconductor device in the direction of line AA. [Figure 3] This is a schematic process diagram illustrating the formation of nanowire channels. [Figure 4] This is a schematic process diagram illustrating the formation of nanowire channels. [Figure 5] This is a schematic process diagram illustrating the formation of nanowire channels. [Figure 6] This is a schematic process diagram illustrating the formation of nanowire channels. [Figure 7] This is a schematic process diagram illustrating the formation of nanowire channels. [Figure 8] This graph illustrates the relationship between the proportion of CH2F2 in the gas used for isotropic etching and the selectivity ratio of silicon germanium to silicon. [Figure 9] This is the result of analyzing the surface composition of the silicon-containing layer on which the protective film was formed. [Figure 10] This is the result of analyzing the surface composition of the silicon germanium-containing layer on which the protective film was formed. [Figure 11] This graph illustrates the relationship between the thickness of the protective film and its effectiveness in preventing the removal of the silicon-containing layer. [Figure 12] This graph illustrates the relationship between the processing time for isotropic etching, the amount of silicon-containing layers and silicon-germanium-containing layers removed, and the selectivity ratio of silicon-germanium to silicon. [Figure 13]A graph for exemplifying the relationship between the number of processing cycles, the removal amount of a layer containing silicon and a layer containing silicon germanium, and the selectivity of silicon germanium with respect to silicon. [Figure 14] A schematic diagram for exemplifying a plasma processing apparatus.
Embodiments for Carrying Out the Invention
[0012] Hereinafter, embodiments will be exemplified while referring to the drawings. In each drawing, the same components are denoted by the same reference numerals, and detailed descriptions thereof are omitted as appropriate. (Method for Manufacturing a Semiconductor Device) The method for manufacturing a semiconductor device according to an embodiment of the present invention can be applied when laminating layers having different constituent elements and anisotropically etching a layer containing one of the elements. Here, as an example, the case where the method for manufacturing a semiconductor device according to an embodiment of the present invention is applied to the manufacture of a MOSFET having a GAA structure will be described.
[0013] In addition, the arrows X, Y, and Z in each of the drawings exemplified below represent three mutually orthogonal directions. For example, the X direction and the Y direction can be directions parallel to the surface of the substrate on which the elements of the semiconductor device are laminated. The Z direction can be the lamination direction of the elements of the semiconductor device.
[0014] FIG. 1 is a schematic plan view of a semiconductor device 300. FIG. 2 is a schematic cross-sectional view taken along the line A-A of the semiconductor device 300 in FIG. 1. As shown in FIGS. 1 and 2, the semiconductor device 300 includes, for example, a substrate 301, a pair of source / drain regions 302, a gate electrode 303, a pair of contact portions 304, and a plurality of nanowire channels 305. That is, the semiconductor device 300 is a MOSFET having a GAA structure.
[0015] The gate electrode 303 is provided between a pair of source and drain regions 302. The contact portion 304 is provided on each of the pair of source and drain regions 302. As shown in Figure 2, multiple nanowire channels 305 are arranged in the Z direction at predetermined intervals. A gate insulating film 306 is provided on the surface of each nanowire channel 305. The nanowire channels 305 with the gate insulating film 306 are covered by a gate electrode 303. An insulating film can also be provided on top of the gate electrode 303. Since known technologies can be applied to the configuration and materials of MOSFETs with a GAA structure, a detailed explanation will be omitted.
[0016] Furthermore, as described above, the semiconductor device manufacturing method according to the embodiment of the present invention can be applied when stacking layers with different elemental compositions and isotropically etching the layer containing one of the elements. Therefore, the semiconductor device manufacturing method according to the embodiment of the present invention can be applied when forming a plurality of nanowire channels 305 arranged in the Z direction at predetermined intervals. In addition, known techniques can be applied to the formation of elements other than the plurality of nanowire channels 305. Therefore, in the following, the method for forming the plurality of nanowire channels 305 will be described, and the method for forming other elements will not be described.
[0017] Figures 3 to 7 are schematic process diagrams illustrating the formation of the nanowire channel 305. Figures 3 to 7 show the state before the semiconductor devices 300 are separated into individual pieces, in the case where multiple semiconductor devices 300 are formed on a single substrate 301 at once. First, multiple layers of silicon germanium-containing layers 307 (corresponding to an example of the first layer) and silicon-containing layers 305a (corresponding to an example of the second layer) are alternately stacked on the substrate 301.
[0018] Next, as shown in Figure 3, the silicon-containing layer 305a and the silicon-germanium-containing layer 307 are sequentially etched using plasma-based anisotropic etching (e.g., RIE: Reactive Ion Etching). As the silicon-containing layer 305a and the silicon-germanium-containing layer 307 are sequentially etched, a laminate 308 is formed in which the silicon-germanium-containing layer 307 and the silicon-containing layer 305a are alternately stacked. In this case, the silicon-containing layer 305a contained in the laminate 308 becomes the nanowire channel 305. Furthermore, since known techniques can be applied to the deposition of the silicon-germanium layer 307 and the silicon-containing layer 305a, and to the formation of the laminate 308, explanations regarding these will be omitted.
[0019] Here, by removing the silicon-germanium-containing layer 307 in the laminate 308, the silicon-containing layer 305a in the laminate 308 can be converted into a nanowire channel 305. In this case, the removal of the silicon-germanium-containing layer 307 can be performed, for example, by isotropic etching using plasma.
[0020] In isotropic etching using plasma, for example, if a gas containing fluorine atoms is used, fluorine radicals are generated. When fluorine radicals come into contact with silicon germanium, the volatile substances SiF4 and GeF4 are generated. Therefore, when performing isotropic etching using plasma, layer 307 containing silicon germanium can be removed by using a gas containing fluorine atoms.
[0021] However, in plasma treatment using a gas containing fluorine atoms, the Si-F bond is more stable than the Ge-F bond. Therefore, silicon is preferentially removed. In other words, simply using a gas containing fluorine atoms in plasma treatment reduces the selectivity ratio of silicon germanium to silicon.
[0022] When the selectivity ratio of silicon-germanium to silicon decreases, a portion of the silicon-containing layer 305a is more likely to be removed when removing the silicon-germanium-containing layer 307. If a portion of the silicon-containing layer 305a is removed, it becomes difficult to form the silicon-containing layer 305a to the desired dimensions. If the silicon-containing layer 305a cannot be formed to the desired dimensions, the function and reliability of the semiconductor device 300 may be reduced.
[0023] In this case, using a gas containing fluorine and hydrogen atoms (corresponding to an example of the second gas), or a gas containing fluorine atoms and hydrogen gas (these gases also correspond to examples of the second gas), fluorine radicals and hydrogen radicals (these radicals correspond to the second radical) are generated. As mentioned above, when fluorine radicals come into contact with silicon germanium, the volatile substances SiF4 and GeF4 are generated. When hydrogen radicals come into contact with silicon germanium, the volatile substance GeH4 is generated. In this case, hydrogen radicals do not react with silicon.
[0024] Therefore, by using a gas containing fluorine atoms and hydrogen atoms, or a gas containing fluorine atoms and hydrogen gas, the selectivity ratio of silicon germanium to silicon can be increased. If the selectivity ratio of silicon germanium to silicon can be increased, the dimensional accuracy and shape accuracy of the formed nanowire channel 305 can be improved. In other words, it becomes possible to form the silicon-containing layer (nanowire channel) to the desired dimensions.
[0025] Gases containing fluorine and hydrogen atoms can be, for example, CH2F2 or HF. Gases containing fluorine atoms and hydrogen gas can be, for example, CF4 and H2. In isotropic etching using plasma, a forming gas may be added to extend the lifetime of radicals. The forming gas is a mixture of nitrogen and hydrogen gas. Adding a forming gas can suppress the recombination and disappearance of generated radicals.
[0026] Figure 8 is a graph illustrating the relationship between the proportion of CH2F2 in the gas used for isotropic etching and the selectivity ratio of silicon germanium to silicon. In addition to CH2F2, the other gases used in isotropic etching are oxygen and a forming gas.
[0027] As mentioned above, increasing the proportion of CH2F2 increases the amount of fluorine radicals, thus increasing both the etching rate of silicon germanium and the etching rate of silicon. In addition, increasing the proportion of CH2F2 increases the amount of hydrogen radicals, which also increases the etching rate of silicon germanium.
[0028] Therefore, as can be seen from Figure 8, increasing the proportion of CH2F2 results in a gradual increase in the etching rate of silicon, while the etching rate of silicon-germanium increases rapidly. In other words, the selectivity ratio of silicon-germanium to silicon can be increased. Furthermore, the same applies when the gas used for isotropic etching is a gas containing fluorine atoms and hydrogen atoms, or when a gas containing fluorine atoms and hydrogen gas are used.
[0029] However, it has been found that if the ratio of gas containing fluorine atoms to gas containing hydrogen atoms, or the ratio of gas containing fluorine atoms to hydrogen gas, is too high in relation to the gas used for isotropic etching, the amount of by-products (deposits) adhering to the inner wall of the chamber where isotropic etching is performed increases. If the amount of by-products adhering to the inner wall of the chamber increases, problems such as large variations in the in-plane distribution of the etching rate may occur. Therefore, it is preferable to keep the ratio of gas containing fluorine atoms to gas containing hydrogen atoms, or the ratio of gas containing fluorine atoms to hydrogen gas, in relation to the gas used for isotropic etching at 60% or less.
[0030] Furthermore, it was found that increasing the ratio of gases containing fluorine atoms and hydrogen atoms, or gases containing fluorine atoms and hydrogen gases, to the gas used for isotropic etching causes the surface of the formed nanowire channels 305 to become rough. This rough surface of the nanowire channels 305 may reduce the functionality and reliability of the semiconductor device 300.
[0031] Therefore, in the method for manufacturing a semiconductor device according to an embodiment of the present invention, a protective film 309 is formed on the surface of the laminate 308 before removing the layer 307 containing silicon germanium. Specifically, a protective film 309 is formed to protect the surface of the silicon-containing layer 305a. If a protective film 309 is formed on the surface of the silicon-containing layer 305a, the dimensional and shape accuracy of the nanowire channel 305 can be improved, and surface roughening of the nanowire channel 305 can be suppressed, even if the ratio of gas containing fluorine atoms and hydrogen atoms, or the ratio of gas containing fluorine atoms and hydrogen gas, to the gas used for isotropic etching is increased in order to increase the selectivity ratio of silicon germanium to silicon.
[0032] For example, after the laminate 308 shown in Figure 3 is formed, as shown in Figure 4, the exposed portion of the laminate 308 is oxidized using an oxygen radical (corresponding to an example of the first radical) or nitrided using a nitrogen radical (corresponding to an example of the first radical) to form a protective film 309 on the surface of the laminate 308. Note that the "surface of the laminate 308" is the exposed portion of the laminate 308 that comes into contact with the oxygen radical or nitrogen radical. In other words, all exposed surfaces of the laminate 308 are included in the "surface of the laminate 308". Therefore, as shown in Figure 4, the "surface of the laminate 308" also includes the sides of the laminate 308.
[0033] For example, by performing isotropic plasma treatment using oxygen gas (corresponding to an example of the first gas) and a forming gas, a protective film 309 containing an oxide can be formed on the surface of the laminate 308. The forming gas can be, for example, a mixture of nitrogen gas and hydrogen gas.
[0034] For example, by performing isotropic plasma treatment using nitrogen gas (corresponding to an example of the first gas), a protective film 309 containing nitride can be formed on the surface of the laminate 308. In this case, hydrogen gas can be added to the nitrogen gas.
[0035] Furthermore, by adding hydrogen gas, the hydrogen radicals react with the native oxide film on the surface of the laminate 308, causing the native oxide film to be etched. film The surface of the etched laminate 308 becomes active, thus promoting the nitriding reaction.
[0036] Furthermore, the formation of the protective film 309 and isotropic etching (removal of the silicon germanium layer 307) can be performed by plasma treatment. Therefore, for example, by switching the gas used, the formation of the protective film 309 and isotropic etching (removal of the silicon germanium layer 307) can be performed sequentially within the same chamber.
[0037] Figure 9 shows the results of analyzing the surface composition of the silicon-containing layer 305a on which the protective film 309 is formed. Note that Figure 9 shows the case where the protective film 309 is an oxide film. As can be seen in Figure 9, by performing a formation process (isotropic plasma treatment using oxygen gas and forming gas) to form a protective film 309 on the surface of the laminate 308, an oxygen-containing film (protective film 309) can be formed on the surface of the silicon-containing layer 305a. Furthermore, by performing an isotropic plasma treatment using a mixed gas of nitrogen gas and hydrogen gas, a nitrogen-containing film (protective film 309) can be formed on the surface of the silicon-containing layer 305a.
[0038] Figure 10 shows the results of analyzing the surface composition of the silicon germanium-containing layer 307 on which the protective film 309 is formed. Note that Figure 10 shows the case where the protective film 309 is an oxide film. As can be seen in Figure 10, by performing a formation process (isotropic plasma treatment using oxygen gas and forming gas) to form a protective film 309 on the surface of the laminate 308, an oxygen-containing film (protective film 309) can be formed on the surface of the silicon germanium-containing layer 307. Furthermore, by performing isotropic plasma treatment using a mixed gas of nitrogen gas and hydrogen gas, a nitrogen-containing film (protective film 309) can be formed on the surface of the silicon germanium-containing layer 307.
[0039] Here, the composition of the portion of the protective film 309 formed on the surface of the silicon germanium-containing layer 307 is as follows: Ta The composition of the portion will be different. Therefore, the etching rate of the portion of the protective film 309 formed on the surface of the silicon germanium-containing layer 307 will be different from that of the portion of the protective film 309 formed on the surface of the silicon-containing layer 305a. Ta The etching rate will differ depending on the area.
[0040] In this case, if isotropic etching is performed using a gas containing fluorine atoms and hydrogen atoms, or a gas containing fluorine atoms and hydrogen gas, the etching rate of the portion of the protective film 309 formed on the surface of the silicon germanium-containing layer 307 will be the same as the etching rate of the portion formed on the surface of the silicon-containing layer 305a of the protective film 309. Ta It was found that the etching rate was higher than that of the partial etching.
[0041] Therefore, when the laminate 308 with a protective film 309 formed on its surface is isotropically etched, as shown in Figure 5, the protective film 309 formed on the surface of the silicon germanium-containing layer 307 is removed first, and then the removal of the silicon germanium-containing layer 307 begins. In this case, if the protective film 309 formed on the surface of the silicon-containing layer 305a remains, it is possible to suppress the removal of the silicon-containing layer 305a or the roughening of the surface of the silicon-containing layer 305a. Note that isotropically etching the laminate 308 with a protective film 309 formed on its surface is sometimes called the process of isotropically etching the laminate 308 with a protective film 309 formed on it.
[0042] However, if the protective film 309 formed on the surface of the silicon-containing layer 305a is removed before the removal of the silicon-germanium-containing layer 307 is completed, the silicon-containing layer 305a may be removed or its surface may become rough. Therefore, the thickness of the protective film 309 needs to be somewhat thick.
[0043] Figure 11 is a graph illustrating the relationship between the thickness of the protective film 309 and its effect in suppressing the removal of the silicon-containing layer 305a. The amount of silicon-containing layer 305a removed is the amount removed when the isotropic etching time for the formed protective film 309 is kept constant. Furthermore, increasing the processing time for forming the protective film 309 allows for a thicker protective film 309 formed on the surface of the silicon-containing layer 305a. The amount of silicon-containing layer 305a removed during processing times from 0 to 60 seconds is inversely proportional to the processing time for forming the protective film 309. In other words, increasing the thickness of the protective film 309 reduces the amount of silicon-containing layer 305a removed.
[0044] As can be seen from Figure 11, if the processing time for forming the protective film 309 is set to 60 seconds or more, the removal of the silicon-containing layer 305a can be almost completely eliminated. However, even if the processing time for forming the protective film 309 is set to 60 seconds or more, it is not possible to completely eliminate the removal of the silicon-containing layer 305a. This is because the protective film 309 can only be formed up to a certain thickness. The reason why a protective film 309 exceeding a certain thickness cannot be formed is thought to be that the distance that radicals can penetrate into the silicon-containing layer 305a is only a few nanometers from the surface of the silicon-containing layer 305a. The reason why radicals can only penetrate a few nanometers from the surface of the silicon-containing layer 305a is thought to be that the lifetime of the radicals ends when they have penetrated a few nanometers from the surface of the silicon-containing layer 305a. In addition, radicals penetrate into the interior of the silicon-containing layer 305a by repeatedly colliding with atoms on the surface of the silicon-containing layer 305a. Therefore, it is thought that energy is lost during collisions. In other words, it is thought that the radicals lose the energy necessary to cause a chemical reaction during collisions with the silicon-containing layer 305a until they penetrate a few nanometers from the surface of the silicon-containing layer 305a. For this reason, it is preferable that the processing time for forming the protective film 309 be about 60 seconds. In this way, the removal of the silicon-containing layer 305a can be almost eliminated, and the decrease in productivity can be suppressed.
[0045] Here, if the thickness of the protective film 309 is made too thin, or if the processing time for removing the silicon germanium-containing layer 307 (isotropic etching processing time) is made too long, the protective film 309 that was formed on the surface of the silicon-containing layer 305a may be removed while the silicon germanium-containing layer 307 is being removed. Furthermore, for example, if the length of the silicon-germanium-containing layer 307 in the Y direction is long, the protective film 309 formed on the surface of the silicon-containing layer 305a may be removed before the silicon-germanium-containing layer 307 is removed.
[0046] In this case, increasing the thickness of the protective film 309 can suppress the exposure of the silicon-containing layer 305a when the silicon-germanium-containing layer 307 is removed. However, increasing the thickness of the protective film 309 reduces productivity, as mentioned above.
[0047] Furthermore, as the removal of the silicon-germanium layer 307 progresses, a portion of the surface 305b of the silicon-containing layer 305a in the Z direction will be exposed, as shown in Figure 5. When a portion of the surface 305b of the silicon-containing layer 305a is exposed, the exposed portion may be removed or the surface of the exposed portion may become rough.
[0048] Figure 12 is a graph illustrating the relationship between the processing time for isotropic etching, the amount of silicon-containing layer 305a and silicon-germanium-containing layer 307 removed, and the selectivity ratio of silicon-germanium to silicon. Figure 12 shows the case where the proportion of CH2F2 in the gas used for isotropic etching is 37%, the processing time for forming the protective film 309 is 60 seconds, and the gas used to form the protective film 309 is oxygen gas and a forming gas (a mixture of nitrogen gas and hydrogen gas). Furthermore, the proportion of CH2F2 was set to 37%, which is lower than the etching rate when the proportion of CH2F2 is 60%, in order to improve the etching accuracy.
[0049] As can be seen from Figure 12, as the isotropic etching process time increases, the protective film 309 formed on the surface of the silicon-containing layer 305a is removed, exposing the silicon-containing layer 305a. As a result, the silicon-containing layer 305a is removed. When the silicon-containing layer 305a is removed, the selectivity ratio of silicon germanium to silicon decreases. This means that if the silicon germanium-containing layer 307 is to be removed by an amount greater than a certain amount (for example, more than 20 nm), it will not be possible to maintain a high selectivity ratio of silicon germanium to silicon.
[0050] Therefore, if necessary, the protective film 309 can be re-formed as shown in Figure 6. The re-formation of the protective film 309 can be carried out in the same manner as when the protective film 309 is formed on the surface of the laminate 308 described above. In this case, for example, by switching the gas used, isotropic etching (removal of the silicon germanium layer 307) and the re-formation of the protective film 309 can be carried out sequentially inside the same chamber.
[0051] Furthermore, for example, if the length of the silicon-germanium-containing layer 307 in the Y direction is short, the removal of the silicon-germanium-containing layer 307 may be completed before the protective film 309 formed on the surface of the silicon-containing layer 305a is removed. Therefore, the formation of the protective film 309 only needs to be performed at least once.
[0052] However, even if the length of the silicon germanium-containing layer 307 in the Y direction is short, as described above, as the removal of the silicon germanium-containing layer 307 progresses, a portion of the surface 305b of the silicon-containing layer 305a will be exposed. For this reason, it is preferable to form the protective film 309 multiple times. The number of times the protective film 309 is formed can be appropriately changed depending on the thickness of the protective film 309 and the length of the silicon germanium-containing layer 307 in the Y direction. For example, the number of times the protective film 309 is formed can be appropriately determined by, for example, conducting experiments or simulations.
[0053] Furthermore, the timing for reforming the protective film 309 can be controlled, for example, by the time elapsed since the start of isotropic etching (removal of the silicon germanium-containing layer 307). In this case, the timing for reforming the protective film 309 can be appropriately determined, for example, by conducting experiments or simulations.
[0054] Furthermore, when the protective film 309 formed on the surface of the silicon-containing layer 305a is removed, the silicon-containing layer 305a is exposed. When the silicon-containing layer 305a is exposed, the emission spectrum in isotropic etching changes. Therefore, by detecting the change in the emission spectrum, it is possible to determine the timing for reforming the protective film 309.
[0055] Figure 13 is a graph illustrating the relationship between the number of processing cycles, the amount of silicon-containing layer 305a and silicon-germanium-containing layer 307 removed, and the selectivity ratio of silicon-germanium to silicon.
[0056] Figure 13 shows the case where the proportion of CH2F2 in the gas used for isotropic etching is 37%, the processing time for forming the protective film 309 is 60 seconds, and the gas used to form the protective film 309 is oxygen gas and a forming gas (a mixture of nitrogen gas and hydrogen gas). Furthermore, the number of processing cycles is defined as "formation of protective film 309 and isotropic etching" being counted as one cycle. For example, if the number of cycles is 4, the "formation of protective film 309 and isotropic etching" is performed four times consecutively. Furthermore, the total time for isotropic etching is set to 600 seconds regardless of the number of cycles. For example, if the number of cycles is 1, isotropic etching is performed once, and the time for isotropic etching is 600 seconds. For example, if the number of cycles is 4, isotropic etching is performed 4 times, and the time for each isotropic etching is 150 seconds (total time for isotropic etching: the total time for 4 isotropic etchings is 600 seconds).
[0057] As can be seen from Figure 13, increasing the number of cycles allows for a higher selectivity ratio of silicon-germanium to silicon. By increasing the selectivity ratio of silicon-germanium to silicon, the amount of material removed from the silicon-containing layer 305a (nanowire channel 305) can be reduced. Furthermore, when isotropically etching the silicon germanium-containing layer 307 in the laminate 308, the silicon germanium-containing layer 307 is etched isotropically from the X and Y directions. Therefore, there is a risk that the amount of material removed from the silicon germanium-containing layer 307 may vary in the etched region. However, if a protective film 309 is formed on the surface of the silicon-containing layer 305a, it is possible to remove only the remaining silicon germanium-containing layer 307 without removing much of the silicon-containing layer 305a. In the method for forming the protective film 309 according to the embodiment of the present invention, a protective film 309 of a constant thickness can be formed by setting the processing time for forming the protective film 309 to 60 seconds or more. In other words, even if there is variation in the amount of material removed from the silicon germanium-containing layer 307 in the etched region, it is possible to suppress variations in the selectivity ratio of silicon germanium to silicon in the laminate 308 in the etched region. If the selectivity ratio of silicon germanium to silicon in the laminate 308 can be suppressed in the etching region, then variations in the dimensions of the silicon-containing layers (nanowire channels) in the laminate can be suppressed. In other words, the dimensional accuracy and shape accuracy of the silicon-containing layer 305a (nanowire channel 305) can be improved, and surface roughness of the silicon-containing layer 305a (nanowire channel 305) can be suppressed. As a result, the functionality and reliability of the semiconductor device 300 can be further improved. Furthermore, as mentioned above, it is preferable that the ratio of gas containing fluorine atoms to hydrogen gas be 60% or less. In particular, when the processing cycle is performed multiple times, it is preferable that the ratio of gas containing fluorine atoms to hydrogen gas be between 25% and 37%. By keeping the ratio of gas containing fluorine atoms to hydrogen gas within the above range, the etching rate of silicon becomes almost zero. Therefore, even if the protective film 309 is etched away, it is possible to suppress the etching of the silicon-containing layer 305a.
[0058] As described above, once the removal of the silicon-germanium layer 307 is complete, the protective film 309 remaining on the surface of the silicon-containing layer 305a is removed, as shown in Figure 7. For example, by continuing isotropic etching even after the removal of the silicon-germanium layer 307 is complete, the protective film 309 remaining on the surface of the silicon-containing layer 305a can be removed. The endpoint of isotropic etching can be determined by time management or changes in the emission spectrum.
[0059] Furthermore, removing the protective film 309 remaining on the surface of the silicon-containing layer 305a may reduce the dimensional and shape accuracy of the silicon-containing layer 305a (nanowire channel 305), or cause the surface of the silicon-containing layer 305a (nanowire channel 305) to become rough. For this reason, once the removal of the silicon-germanium-containing layer 307 is complete, a protective film 309 covering the surface of the silicon-containing layer 305a may be formed again. The formed protective film 309 can be, for example, the gate insulating film 306 shown in Figure 2. Note that an insulating film different from the protective film 309 may also be formed.
[0060] As described above, the method for manufacturing a semiconductor device according to an embodiment of the present invention may include the following steps. A step of forming a protective film 309 on the surface of a laminate 308 in which a first layer (307) and a second layer (305a) having a different composition of elements from the first layer (307) are stacked. A process of isotropically etching a laminate on which a protective film has been formed. The process of forming the protective film 309 is carried out using a first radical generated by exciting and activating the first gas G1 with plasma P. The process of isotropically etching the laminate 308 is carried out using a second radical generated by exciting and activating a second gas G2, which is different from the first gas G1, with plasma P. The steps of forming the protective film 309 and isotropically etching the laminate 308 are carried out sequentially in the same atmosphere.
[0061] In the process of forming the protective film 309, the etching rate of the portion of the protective film 309 formed on the surface of the first layer (307) is higher than the etching rate of the portion of the protective film 309 formed on the surface of the second layer (305a).
[0062] In the process of isotropically etching the laminate 308, if it is determined that the protective film 309 formed on the surface of the second layer (305a) has been removed, the isotropic etching is stopped and the process of forming the protective film 309 is performed.
[0063] The removal of the protective film 309 is determined by at least one of the following: the time elapsed since the start of isotropic etching, and the change in the emission spectrum.
[0064] The steps of forming the protective film 309 and isotropically etching the laminate 308 are performed alternately multiple times.
[0065] (Plasma treatment device) The plasma processing apparatus according to the embodiment of the present invention is only required to be capable of forming the nanowire channels 305 described above. For example, the plasma processing apparatus can generate plasma, excite and activate a gas with the generated plasma to produce radicals and ions, and mainly perform processing using the generated radicals. By mainly performing processing using radicals, isotropic processing can be performed.
[0066] Figure 14 is a schematic diagram illustrating the plasma processing apparatus 1. Plasma processing apparatus 1 is a microwave-excited plasma processing apparatus also known as a CDE (chemical dry etching) apparatus or a remote plasma apparatus.
[0067] As shown in Figure 14, the plasma processing apparatus 1 includes, for example, a chamber 2, a mounting section 3, a depressurization section 4, a discharge tube 5, a microwave generation section 6, a gas supply section 7, a detection section 8, and a controller 9.
[0068] Chamber 2 has an airtight structure capable of maintaining an atmosphere reduced to atmospheric pressure. A flow straightening plate 2c can be provided inside Chamber 2. The flow straightening plate 2c is plate-shaped and has multiple holes that penetrate in the thickness direction.
[0069] The mounting section 3 is provided inside the chamber 2 and is used to place the workpiece 100. The workpiece 100 is, for example, a substrate 301 on which the aforementioned laminate 308 is formed. The mounting section 3 is provided on the bottom surface of the chamber 2.
[0070] The pressure reduction unit 4 reduces the pressure inside the chamber 2 so that the atmosphere inside the chamber 2 reaches a predetermined pressure. The pressure reduction unit 4 is connected via piping to an exhaust port 2a provided in the chamber 2. The pressure reduction unit 4 includes, for example, a pump and a pressure control unit. The pump is, for example, a turbomolecular pump (TMP). The pressure control unit is, for example, an APC (Auto Pressure Controller).
[0071] The discharge tube 5 is located outside the chamber 2. The discharge tube 5 is tubular in shape, and one end is connected to an inlet hole 2b provided in the ceiling of the chamber 2 via a piping member 4a. The discharge tube 5 is made of a dielectric material. The discharge tube 5 can be made of, for example, quartz.
[0072] The microwave generation unit 6 includes a microwave source 6a and a waveguide 6b. The microwave source 6a generates microwaves M at a predetermined frequency (e.g., 2.75 GHz) and radiates them toward the waveguide 6b. One end of the waveguide 6b is connected to the microwave source 6a. The other end of the waveguide 6b is connected to the discharge tube 5. The waveguide 6b extends in a direction intersecting the direction in which the discharge tube 5 extends. The microwaves M propagate inside the waveguide 6b and are introduced into the discharge tube 5 through a slot provided at the end of the waveguide 6b. The internal space 5a of the discharge tube 5, near the location of the slot in the waveguide 6b, becomes the region where plasma P is generated.
[0073] The gas supply unit 7 is connected to the end of the discharge tube 5 opposite to the chamber 2 side. The gas supply unit 7 switches between supplying gas G1 (corresponding to an example of the first gas) used for forming the protective film 309 and gas G2 (corresponding to an example of the second gas) used for isotropic etching (removal of the silicon germanium layer 307) into the discharge tube 5.
[0074] For example, the gas supply unit 7 includes a storage unit 7a for storing gas G1, a control unit 7b for controlling gas G1, a storage unit 7c for storing gas G2, and a control unit 7d for controlling gas G2. The storage units 7a and 7c can be, for example, high-pressure cylinders. The control units 7b and 7d can have the function of switching between supplying and stopping the supply of gas, and the function of controlling the gas flow rate. The storage unit 7a and the control unit 7b constitute a first gas supply unit that supplies gas G1 to the region where plasma P is generated. The storage unit 7c and the control unit 7d constitute a second gas supply unit that supplies gas G2 to the region where plasma P is generated. Furthermore, a storage unit and control unit for supplying the aforementioned foaming gas and hydrogen gas can also be provided.
[0075] Gases G1 and G2 supplied into the discharge tube 5 are excited and activated by the plasma, generating plasma products such as radicals, ions, and electrons. These generated plasma products are supplied from the discharge tube 5 to the chamber 2 via the piping member 4a. In this process, short-lived ions and electrons cannot reach the inside of the chamber 2, and radicals are supplied to the inside of the chamber 2. When the radicals reach the workpiece 100, the surface of the workpiece 100 is chemically treated. That is, an isotropic treatment mainly consisting of radicals is performed on the workpiece 100.
[0076] The detection unit 8 is located outside the chamber 2. The detection unit 8 detects light emission near the processing surface of the workpiece 100 through a light-transmitting window 2d provided on the wall of the chamber 2. The detected light emission data is sent from the detection unit 8 to the controller 9. The light emission data can be used to detect changes in the emission spectrum (endpoint detection), that is, to determine the timing for reforming the protective film 309 described above.
[0077] The controller 9 comprises an arithmetic unit such as a CPU (Central Processing Unit) and a storage unit such as memory. The controller 9 is, for example, a computer. The controller 9 controls the operation of each element provided in the plasma processing apparatus 1 based on a control program stored in the storage unit.
[0078] Known techniques can be applied to procedures such as the generation procedure and conditions for plasma P, the generation of plasma products such as radicals, isotropic treatment using radical cal, and the loading and unloading of the workpiece 100. Therefore, the following will explain the switching between gas G1 and gas G2 (the switching between the formation of the protective film 309 and isotropic etching (removal of the silicon germanium layer 307)).
[0079] For example, the controller 9 detects a change in the emission spectrum based on the detected value from the detection unit 8, and if it determines that the protective film 309 formed on the surface of the silicon-containing layer 305a has been removed, it stops isotropic etching (removal of the silicon-germanium-containing layer 307) and reforms the protective film 309. For example, the controller 9 controls the control unit 7d of the gas supply unit 7 to stop the supply of gas G2 and controls the control unit 7b to supply gas G1. As mentioned above, the timing of reforming the protective film 309 can also be based on a predetermined time. However, reforming the protective film 309 based on a change in the emission spectrum makes it possible to handle variations in process conditions. Furthermore, the controller 9 can also alternately perform the formation of the protective film 309 and the isotropic etching of the laminate 308 multiple times.
[0080] Although plasma processing apparatus 1 was described as an example, any plasma processing apparatus that primarily uses radicals for processing is acceptable. Furthermore, while the example described is the case in which multiple semiconductor devices 300 are formed on a single substrate 301 at once, the present invention also includes cases where the devices are not separated into individual pieces. For example, the present invention is also effective in the case of a semiconductor device in which nanowire channels 305 are arranged at predetermined intervals not only in the Z direction but also in the Y direction.
[0081] The embodiments have been illustrated above. However, the present invention is not limited to these descriptions. For example, any embodiments described above in which a person skilled in the art has added, deleted, or modified components, or added, omitted, or modified processes, are also included within the scope of the present invention, as long as they retain the features of the present invention. For example, the shape, dimensions, material, and arrangement of each element of the plasma processing apparatus 1 are not limited to those exemplified and can be changed as appropriate. Furthermore, the elements of each embodiment described above can be combined as much as possible, and these combinations are also included within the scope of the present invention insofar as they include the features of the present invention. [Explanation of symbols]
[0082] 1 Plasma processing apparatus, 5 Discharge tube, 6 Microwave generation unit, 7 Gas supply unit, 8 Detection unit, 9 Controller, 100 Processing material, 300 Semiconductor device, 301 Substrate, 305 Nanowire channel, 305a Silicon-containing layer, 305b Surface, 307 Silicon-germanium-containing layer, 308 Laminate, 309 Protective film, G2 Gas, G1 Gas, M Microwave, P Plasma
Claims
1. A step of forming a protective film on the surface including the side surface of a laminate in which a first layer and a second layer having different elemental composition from the first layer are laminated, A step of isotropically etching the laminate on which the protective film is formed, Equipped with, The process of forming the protective film is carried out using a first radical generated by exciting and activating a first gas with plasma. The process of isotropically etching the laminate is carried out using a second radical generated by exciting and activating a second gas, different from the first gas, with plasma. In the process of isotropically etching the laminate, if it is determined that the protective film formed on the surface of the second layer has been removed, the isotropic etching is stopped and the process of forming the protective film is performed. A method for manufacturing a semiconductor device, wherein the step of forming the protective film and the step of isotropically etching the laminate are performed in the same atmosphere.
2. A method for manufacturing a semiconductor device according to claim 1, wherein in the step of forming the protective film, the etching rate of the portion of the protective film formed on the surface of the first layer is higher than the etching rate of the portion of the protective film formed on the surface of the second layer.
3. The method for manufacturing a semiconductor device according to claim 1 or 2, wherein the determination that the protective film has been removed is made by at least one of the time elapsed since the start of the isotropic etching and the change in the emission spectrum.
4. A method for manufacturing a semiconductor device according to claim 1 or 2, wherein the step of forming the protective film and the step of isotropically etching the laminate are performed alternately multiple times.
5. The first layer contains silicon germanium, The second layer contains silicon, The first gas is a gas containing at least one of oxygen atoms and nitrogen atoms, The method for manufacturing a semiconductor device according to claim 1 or 2, wherein the second gas is a gas containing fluorine atoms and hydrogen atoms, or a gas containing fluorine atoms and hydrogen gas.
6. A chamber capable of maintaining an atmosphere reduced to below atmospheric pressure, A mounting section is provided inside the chamber for placing a workpiece having a laminate in which a first layer and a second layer having different elemental composition from the first layer are stacked, A plasma generating unit that generates plasma, A first gas supply unit that supplies a first gas to the region where the plasma is generated, A second gas supply unit that supplies a second gas different from the first gas to the region where the plasma is generated, A controller that controls the plasma generation unit, the first gas supply unit, and the second gas supply unit, Equipped with, The aforementioned controller, The plasma generation unit and the first gas supply unit are controlled to excite and activate the first gas with the plasma to generate a first radical, and a protective film is formed on the surface of the laminate using the first radical inside the chamber. The plasma generation unit and the second gas supply unit are controlled to excite and activate the second gas with the plasma to generate a second radical, and the laminate on which the protective film is formed is isotropically etched using the second radical inside the chamber. A plasma processing apparatus that controls the plasma generation unit, the first gas supply unit, and the second gas supply unit, and when it is determined that the protective film formed on the surface of the second layer has been removed, stops the isotropic etching of the laminate using the second radical and forms the protective film using the first radical.
7. The system further includes a detection unit that detects light emission near the processed surface of the processed object. The plasma processing apparatus according to claim 6, wherein the controller detects a change in the emission spectrum based on the detected value from the detection unit, and when it determines that the protective film formed on the surface of the second layer has been removed, it stops the isotropic etching of the laminate using the second radical and forms the protective film using the first radical.
8. The plasma processing apparatus according to claim 6 or 7, wherein the controller alternately performs the formation of the protective film and the isotropic etching of the laminate multiple times.