The present application claims priority of Korean Patent Application No. 10-2020-0135081, filed on Oct. 19, 2020, which is herein incorporated by reference in its entirety.
The disclosure relates to semiconductor devices, and more specifically, to semiconductor devices and methods for manufacturing the semiconductor devices.
Manufacturing electronic devices, such as semiconductor devices, require a gapfill for a three-dimensional structure or high aspect ratio structure. The gapfill of the high aspect ratio structure is performed in the manufacture of, e.g., vertical semiconductor devices.
According to an embodiment of the disclosure, a method for manufacturing a semiconductor device may comprise forming a gapfill target structure on a semiconductor substrate, the gapfill target structure including a horizontal recess parallel with the semiconductor substrate and having a first surface and a vertical slit extending from the horizontal recess and having a second surface perpendicular to the semiconductor substrate; removing a native oxide from the first surface to form a pre-cleaned first surface; forming, in-situ, a first semiconductor material on the pre-cleaned first surface; and forming a second semiconductor material on the first semiconductor material.
According to an embodiment of the disclosure, a method for manufacturing a semiconductor device may comprise forming a lower level stack on a semiconductor substrate, the lower level stack including a source sacrificial layer and a source layer; forming an alternate stack on the lower level stack, the alternate stack including insulation layers and sacrificial layers; forming a vertical channel structure including a channel layer penetrating the alternate stack and the lower level stack; forming a slit exposing the source sacrificial layer and penetrating the alternate stack; forming a sealing layer on a side wall of the slit; forming a horizontal recess extending from the slit by removing the source sacrificial layer; exposing a portion of the channel layer from the horizontal recess; exposing an exposed surface of the channel layer to a pre-cleaning process of halogen gas; and selectively growing, in-situ, a polysilicon layer on the exposed surface of the channel layer after the pre-cleaning process.
According to an embodiment of the disclosure, a semiconductor device may comprise an alternate stack including insulation layers and gate electrodes alternately stacked one over another, on a semiconductor substrate; a source channel contact layer between the semiconductor substrate and the alternate stack; a vertical channel layer penetrating the alternate stack and the source channel contact layer; and a memory layer between the vertical channel layer and the alternate stack, wherein the source channel contact layer includes an epitaxial polysilicon layer contacting the vertical channel layer; and an amorphous silicon layer on the epitaxial polysilicon layer.
Hereinafter, embodiments of the disclosure are described with reference to schematic cross-sectional views, plan views, or block diagrams. Changes or modifications may be made to the views depending on manufacturing techniques and/or tolerances. Thus, embodiments of the disclosure are not limited to specific types as shown and illustrated herein but may rather encompass changes or modifications resultant from fabricating processes. For example, the regions or areas shown in the drawings may be schematically shown, and their shapes shown are provided merely as examples, rather as limiting the category or scope of the disclosure.
Embodiments of the disclosure may provide a vertical semiconductor device with better reliability and a method for manufacturing the vertical semiconductor device.
According to the present technology, since native oxides on the surface of the channel layer are removed by a cleaning process using halogen gas, high and uniform current may be secured.
According to the present technology, since the polysilicon layer is selectively grown as the source channel contact layer, it may be possible to prevent phosphorus from accumulating at the interface between the channel layer and the source channel contact layer.
Referring to
The alternate stack 120 may include insulation layers 121 and gate electrodes 122 alternately formed with one another. The lowest insulation layer among the insulation layers 121 may be thicker than the other insulation layers. The insulation layers 121 may include silicon oxide, and the gate electrodes 122 may include a metal-base material. The gate electrodes 122 may include tungsten or a stack of titanium nitride and tungsten.
The vertical semiconductor device 100 may further include a vertical channel structure 130 penetrating the alternate stack 120. The vertical channel structure 130 may include a memory layer 131, a channel layer 132, and a core insulation layer 133. The core insulation layer 133 may fill the internal space of the channel layer 132, and the memory layer 131 may surround the outer wall of the channel layer 132. A lower portion of the vertical channel structure 130 may penetrate the lower level stack 110 and make contact with the semiconductor substrate 101. An upper portion of the vertical channel structure 130 may penetrate the alternate stack 120.
The vertical semiconductor device 100 may further include a slit 119 penetrating the alternate stack 120. The slit 119 may be spaced apart from the vertical channel structure 130. A sealing layer 117 may be formed on the side wall of the slit 119. The slit 119 may be shaped as a trench. The sealing layer 117 may cover first ends of the gate electrodes 122. The sealing layer 117 may include silicon oxide, silicon nitride, carbon-containing silicon oxide, or a combination thereof.
The lower level stack 110 is described below.
The lower level stack 110 may include the source layers 111 and 112 and the source channel contact layer 110S between the source layers 111 and 112. The source layers 111 and 112 may include a lower source layer 111 and an upper source layer 112. The lower level stack 110 may further include a horizontal recess 118. The horizontal recess 118 may be defined between the lower source layer 111 and the upper source layer 112. The source channel contact layer 110S may be formed between the lower source layer 111 and the upper source layer 112. The source channel contact layer 110S may fill the horizontal recess 118. The lower source layer 111 and the upper source layer 112 may include the same material, e.g., a semiconductor material, such as polysilicon. The source channel contact layer 110S may include a semiconductor material, e.g., silicon.
The source channel contact layer 110S may include a first silicon layer 113 and a second silicon layer 114. The first silicon layer 113 may cover the surface of the horizontal recess 118. The first silicon layer 113 may directly contact the channel layer 132 of the vertical channel structure 130. A portion of the first silicon layer 113 may extend to cover a bottom portion of the slit 119, contacting the sealing layer 117. The second silicon layer 114 may fill the horizontal recess 118 on the first silicon layer 113 and extend to fill the slit 119.
The first silicon layer 113 and the second silicon layer 114 may be silicon layers having different crystalline phases. The first silicon layer 113 may be a crystalline silicon layer, and the second silicon layer 114 may be an amorphous silicon layer. The first silicon layer 113 may be an epitaxial polysilicon layer, and the second silicon layer 114 may be an amorphous silicon layer. The first silicon layer 113 may be an epitaxial polysilicon layer, and the second silicon layer 114 may be a deposited amorphous silicon layer. The epitaxial polysilicon layer may be formed by epitaxial growth, and the deposited amorphous silicon layer may be formed by deposition. The first silicon layer 113 and the second silicon layer 114 may include a dopant. The dopant may include phosphorus. The first silicon layer 113 may include a phosphorus-doped epitaxial polysilicon layer, and the second silicon layer 114 may include a phosphorus-doped amorphous silicon layer.
The source channel contact layer 110S may further include an interface layer 115 between the first silicon layer 113 and the second silicon layer 114. The interface layer 115 may include silicon oxide. The interface layer 115 may include an oxide of the first silicon layer 113. The interface layer 115 may be thinner than the first silicon layer 113 and the second silicon layer 114. The interface layer 115 may serve to improve the surface roughness of the first silicon layer 113, thereby preventing seams or voids of the second silicon layer 114. The interface layer 115 may be extremely thin for electrical contact between the first silicon layer 113 and the second silicon layer 114.
As described above, the contact surface between the channel layer 132 and the first silicon layer 113 may include an oxide-free surface, and the contact surface between the first silicon layer 113 and the second silicon layer 114 may include an oxidized surface. The oxidized surface may include the interface layer 115. The oxide-free surface refers to a surface in which oxide is not present, and the surface of the channel layer 132 has a pre-cleaned surface, and the first silicon layer 113 may be selectively grown on the pre-cleaned surface.
As shown in
Next, an upper level stack including insulation layers 17 and sacrificial layers 18 may be formed on the upper source layer 16. The upper level stack may include the insulation layers 17 and the sacrificial layers 18 alternately stacked one over another. The insulation layers 17 and the sacrificial layers 18 may be stacked alternately several times. The insulation materials 17 and the sacrificial layers 18 may include different materials. The insulation layers 17 may have etch selectivity to the sacrificial layers 18. The insulation layers 17 may include silicon oxide, and the sacrificial layers 18 may include silicon nitride. The insulation layers 17 and the sacrificial layers 18 may have the same thickness. The insulation layers 17 and the sacrificial layers 18 may be thicker than the liner layers 13 and 15, and the insulation layers 17 and the sacrificial layers 18 may be thinner than the lower source layer 12 and the upper source layer 16. Among the insulation layers 17, the lowest insulation layer 17 may be thicker than the other insulation layers 17.
The insulation layers 17 and the sacrificial layers 18 may be formed using chemical vapor deposition (CVD) or atomic layer deposition (ALD).
Referring to
The vertical opening 19 may be formed to be perpendicular to the surface of the semiconductor substrate 11. The vertical opening 19 may be shaped to penetrate the insulation layers 17 and the sacrificial layers 18 and extend to penetrate the upper source layer 16, the liner layers 13 and 15, the source sacrificial layer 14, and the lower source layer 12. Although not shown, at plan view, a plurality of vertical openings 19 may be formed and may have a hole array structure. Upon forming the vertical opening 19, the surface of the semiconductor substrate 101 may be recessed. According to an embodiment, the vertical opening 19 may be denoted a ‘vertical recess,’ ‘vertical hole,’ or ‘channel hole.’
Referring to
The vertical channel structure 20 may include a memory layer 21, a channel layer 22, and a core insulation layer 23. The memory layer 21 may have a stack structure including a blocking layer, a charge trapping layer, and a tunnel insulation layer. The blocking layer and the tunnel insulation layer may include oxide, and the charge trapping layer may include nitride. The memory layer 21 may have an oxide-nitride-oxide (ONO) structure. The channel layer 22 may include an undoped polysilicon layer without impurities. The channel layer 22 may have a cylinder shape having an inner space. The memory layer 21 may surround the outer wall of the channel layer 22. The internal space of the channel layer 22 may be fully filled with the core insulation layer 23. The core insulation layer 23 may include silicon oxide or silicon nitride.
Referring to
Referring to
Referring to
Next, the upper source layer 16, liner layer 15, and source sacrificial layer 14 may be etched using the sealing layer 26 as a barrier. The liner layer 13 and the lower source layer 12 may be left without being etched out.
Referring to
The horizontal recess 27 may expose a lower side wall of the vertical channel structure 20. The outer wall of the vertical channel structure 20 may be a portion of the memory layer 21. At top view, the horizontal recess 27 may be shaped to surround the lower side wall of the vertical channel structure 20.
Referring to
After the liner layers 13 and 15 are removed, a portion of the memory layer 21 of the vertical channel structure 20 may be removed.
By the above-described series of processes, the horizontal recess 28 may expose the lower outer wall of the channel layer 22. A portion of the memory layer 21 may be cut by the horizontal recess 28. Thus, an undercut 28E may be formed between the channel layer 22 and the lower/upper source layers 12 and 16.
The horizontal recess 28 may be parallel with the semiconductor substrate 11 and have a first surface. The slit 24 may extend from the horizontal recess 28 and have a second surface perpendicular to the semiconductor substrate 11. In other words, a gapfill target structure including the horizontal recess 28 having the first surface and the slit 24 having the second surface may be formed on the semiconductor substrate 11. The first surface may be provided by the channel layer 22, lower source layer 12, and upper source layer 16, and the second surface may be provided by the sealing layer 26. The first surface may be a surface of a silicon layer, and the second surface may be a surface of an insulation material.
Subsequently, by the series of processes shown in
Referring to
The pre-treatment process 29 may include a process for thinning or removing the native oxide remaining on the exposed surface of the channel layer 22. The pre-treatment process 29 may be performed using a fluorine-based chemical. After removing or thinning the native oxide as thin as possible, it needs to subsequently be loaded in the furnace chamber. The time taken from the pre-treatment process 29 to the loading of the substrate in the furnace chamber may be within two hours. The fluorine-based chemical may include NF3 or HF.
Referring to
Next, a pre-cleaning process 30 may be performed, using an in-situ etching gas, in the furnace chamber. The native oxide on the surface of the channel layer 22, which is inevitably formed when moving and loading the substrate, may be removed by the pre-cleaning process 30. The etching gas for the pre-cleaning process 30 may include halogen gas, such as Cl or HBr. The removal of native oxide using the etching gas may be performed in such a manner that the etching gas infiltrates through tiny gaps present in the native oxide and etches the channel layer 22 and then lifts off the native oxide. Thus, if the time taken from the pre-treatment process 29 to the loading of the substrate into the furnace chamber increases, the tiny gaps in the native oxide may vanish, drastically lowering the in-situ etching efficiency. By the pre-cleaning process 30, the native oxide present on the exposed surface of the lower source layer 12 and the upper source layer 16 may be removed as well. The channel layer 22 may include the pre-cleaned surface that may be oxide-free. The exposed surface of the lower source layer 12 and the upper source layer 16 may also include the oxide-free, pre-cleaned surface.
Referring to
As such, the deposition process of the polysilicon layer 31 may selectively epitaxial-grow the polysilicon layer 31 on the exposed surface of the pre-cleaned channel layer 22 while simultaneously forming, selectively, the sacrificial amorphous silicon layer 32 on the surface of the sealing layer 26. The words “simultaneous” and “simultaneously” as used herein with respect to occurrences mean that the occurrences take place on overlapping intervals of time. For example, if a first occurrence takes place over a first interval of time and a second occurrence takes place simultaneously over a second interval of time, then the first and second intervals at least partially overlap each other such that there exists a time at which the first and second occurrences are both taking place.
The deposition process of the polysilicon layer 31 and the sacrificial amorphous silicon layer 32 may adjust the mixing ratio of a chlorine-containing silicon source material to a chlorine-free silicon source material, thereby adjusting the deposition ratio of the polysilicon layer 31 to the sacrificial amorphous silicon layer 32. The chlorine-containing silicon source material may include dichlorosilane (SiH2Cl2, DCS), and the chlorine-free silicon source material may include monosilane (SiH4). The deposition process of the polysilicon layer 31 and the sacrificial amorphous silicon layer 32 may be performed, with the proportion of the chlorine-free silicon source material larger than the proportion of the chlorine-containing silicon source material.
The mixing ratio and pressure of dichlorosilane (DCS) and monosilane (SiH4) may increase the growth rate of the polysilicon layer 31 to a level equal to that of the sacrificial amorphous silicon layer 32. For example, the process temperature may be 450° C. to 490° C., the mixing ratio of monosilane to dichlorosilane may be 7:1 to 9:1, and the pressure may be set to less than 1 Torr.
If the proportion of dichlorosilane (DCS) increases, the formation rate of the polysilicon layer 31 is higher than that of the sacrificial amorphous silicon layer 32, but uniformity within the wafer may deteriorate.
If the proportion of monosilane (SiH4) is increased, the uniformity in the wafer is improved, but the formation rate of the sacrificial amorphous silicon layer 32 is increased, and the inside of the slit 24 may be blocked. If the slit 24 is blocked, the polysilicon layer is difficult to form.
The pressure needs to be less than 1 Torr to secure deposition uniformity in the wafer.
For example, when the ratio of monosilane (SiH4) to dichlorosilane (DCS) is 3:1, the deposition ratio of the polysilicon layer 31 to the sacrificial amorphous silicon layer 32 may be about 1.5:1. In this case, the uniformity is about 5 to 9% (0.5 Torr to 4.5 Torr).
When the ratio of monosilane (SiH4) to dichlorosilane (DCS) is 8:1, the deposition ratio of the polysilicon layer 31 to the sacrificial amorphous silicon layer 32 may be about 1.1:1, and the uniformity is about 2 to 4% (0.5 Torr to 4.5 Torr).
The deposition ratio of the polysilicon layer 31 to the sacrificial amorphous silicon layer 32 according to the gas ratio may be maintained in pressure changes ranging from 0.5 Torr to 4.5 Torr.
The principle of adjusting the deposition ratio of the polysilicon layer 31 to the sacrificial amorphous silicon layer 32 according to the ratio of monosilane (SiH4) to dichlorosilane (DCS) is as follows. Cl2 gas generated from dichlorosilane (DCS) may play a role to suppress formation of the sacrificial amorphous silicon layer 32 (or etching simultaneously with deposition). As the amount of Cl2 gas increases, the deposition rate of the sacrificial amorphous silicon layer 32 decreases, but the deposition of the polysilicon layer 31 might not be suppressed. Resultantly, the difference in thickness between the polysilicon layer 31 and the sacrificial amorphous silicon layer 32 may be adjusted according to the proportion of dichlorosilane (DCS).
Meanwhile, as the proportion of dichlorosilane (DCS) increases, the thickness distribution is deteriorated. Therefore, the ratio of monosilane (SiH4) to dichlorosilane (DCS) may be set to 8:1, improving the thickness distribution. Further, the deposition ratio of the polysilicon layer 31 to the sacrificial amorphous silicon layer 32 may be optimized by setting the ratio of monosilane (SiH4) to dichlorosilane (DCS) to 8:1.
As described above, the polysilicon layer 31 is deposited on the surface of the horizontal recess 28 using a mixed gas of monosilane (SiH4) and dichlorosilane (DCS), and the sacrificial amorphous silicon layer 32 is rendered to be deposited on the sealing layer 26 by optimizing the mixing ratio of monosilane (SiH4) to dichlorosilane (DCS).
A broken silicon lattice may exist on the surfaces of the channel layer 22, the lower source layer 12, and the upper source layer 16 from which the native oxides have been removed by the pre-cleaning process 30 using a halogen gas. Accordingly, since the energy of growth in the crystalline direction is low, the polysilicon layer 31 is epitaxially grown on the surfaces of the channel layer 22, the lower source layer 12 and the upper source layer 16.
The polysilicon layer 31 may be thinner than the sacrificial amorphous silicon layer 32. The sacrificial amorphous silicon layer 32 might not fully fill the inside of the slit 24. The polysilicon layer 31 and the sacrificial amorphous silicon layer 32 may have the same thickness.
In another embodiment, the deposition process of the polysilicon layer 31 and the sacrificial amorphous silicon layer 32 may be performed using mono silane (SiH4) alone.
Referring to
For example, when the sacrificial amorphous silicon layer 32 is grown to about 400 Å and the polysilicon layer 31 is grown to about 250 Å, the polysilicon layer 32 may be etched by 400 Å while etching the sacrificial amorphous silicon layer 32 by 400 Å using HBr gas. Resultantly, after the sacrificial amorphous silicon layer 32 is fully removed, the polysilicon layer 31 may be left with a thickness of 210 Å.
As such, after the sacrificial amorphous silicon layer 32 is fully removed, the polysilicon layer 31 may remain. The polysilicon layer 31 may remain in the horizontal recess 28.
Referring to
Referring to
By the above-described series of processes, the horizontal recess 28 may be void-free and be filled with the polysilicon layer 31, interface layer 33, and amorphous silicon layer 34. The slit 24 may be filled with the amorphous silicon layer 34. A portion of the polysilicon layer 31 may extend to cover a bottom portion of the slit 24, contacting the bottom surface of the sealing layer 26. The amorphous silicon layer 34 may fill the horizontal recess 28 and extend to fill the slit 24. The contact surface between the polysilicon layer 31 and the amorphous silicon layer 34 may include the interface layer 33, and the interface layer 33 may include oxide.
Referring to
First, the horizontal recess 28 may be formed by a series of processes as shown in
Next, as shown in
As such, if the horizontal recess 28 is fully filled with the polysilicon layer 31′, the process is simplified.
The description made above in connection with
The mixing ratio and pressure of dichlorosilane (DCS) and monosilane (SiH4) may increase the growth rate of the polysilicon layer 31′ to a level equal to that of the sacrificial amorphous silicon layer 32′. For example, the process temperature may be 450° C. to 490° C., the mixing ratio of monosilane to dichlorosilane may be 7:1 to 9:1, and the pressure may be set to less than 1 Torr.
As described above, the polysilicon layer 31′ is deposited on the surface of the horizontal recess 28 using a mixed gas of monosilane (SiH4) and dichlorosilane (DCS), and the sacrificial amorphous silicon layer 32′ is rendered to be deposited on the sealing layer 26 by optimizing the mixing ratio of monosilane (SiH4) to dichlorosilane (DCS).
In another embodiment, the deposition process of the polysilicon layer 31′ and the sacrificial amorphous silicon layer 32′ may be performed using mono silane (SiH4) alone.
Referring to
After the sacrificial amorphous silicon layer 32′ is removed, the polysilicon layer 31′ may remain in the horizontal recess 28. The polysilicon layer 31′ may fill the horizontal recess 28.
Referring to
An air gap AG may be formed between the polysilicon layer 31′ and the amorphous silicon layer 34′. Alternatively, the air gap AG may be filled with the amorphous silicon layer 34′.
Referring to
According to the above-described embodiments, since native oxides are removed by the pre-cleaning process 30 using halogen gas, phosphorus (Ph) diffused from the polysilicon layers 31 and 31′ may be easily controlled. Thus, a NAND operation, in particular an erase operation using gate induced drain leakage (GIDL) current may be smoothly performed.
Further, since the native oxides are removed by the pre-cleaning process 30 using halogen gas, current inhibitors between the channel layer 22 and the polysilicon layers 31 and 31′ may be freed, rendering it possible to secure a high and uniform current.
Further, since the polysilicon layers 31 and 31′ and the sacrificial amorphous silicon layers 32 and 32′ are selectively formed, a relatively high etch rate may be secured as compared with the other part of the polysilicon layers 31 and 31′ when a subsequent etching process proceeds. Thus, it is easy to secure an etch margin in the subsequent etching process.
Since it is easy to secure an etching margin, the incidence of defective etching of the polysilicon layers 31 and 31′ due to the etching process may be reduced. For example, use of a combination of the polysilicon layer 31 and the sacrificial amorphous silicon layer 32 may further reduce defects.
Referring to
In another embodiment, a selective polysilicon layer deposition process may be applied to drain contact layers, as well as to the source channel contact layers.
In another embodiment, a selective polysilicon layer deposition process is a low-temperature process (450° C. to 490° C.) as compared with normal epitaxy processes and be applied in low-temperature epitaxy processes.
As another embodiment, since the selective polysilicon layer deposition process is able to adjust the ratio of polysilicon layer to amorphous silicon layer, it is also applicable to hard mask processes of amorphous silicon layer/polysilicon layer that requires a difference in light transmittance. By using the nature that the transmittance of polysilicon layer is higher than the transmittance of amorphous silicon layer, a hard mask may be formed so that areas may be differentiated by partial amorphous and crystalline formation. For example, to form the vertical opening 19 of
While the present invention has been described with respect to specific embodiments, it will be apparent to those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention as defined in the following claims.
| Number | Date | Country | Kind |
|---|---|---|---|
| 10-2020-0135081 | Oct 2020 | KR | national |