Method for producing single-crystal silicon doped with n-type dopants by the CZ method
The method addresses uniform dopant distribution and dislocation prevention in single-crystal silicon by controlling gaseous dopant flow through a pressure-regulated valve system, achieving consistent resistivity and stable growth in the CZ method.
Patent Information
- Application Number
- JP2025521162
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2022-10-14
- Filing Date
- 2023-09-28
- Publication Date
- 2025-10-14
AI Technical Summary
Existing methods face challenges in achieving uniform dopant distribution along the axial length of single-crystal silicon and preventing dislocation formation during the CZ method, particularly with sublimable dopants like phosphorus and arsenic, which have high vapor pressures and require precise control.
A method involving a sublimation unit outside the reactor chamber where a solid dopant is heated to form gaseous dopant, controlled by a pressure difference-based valve system, ensuring a constant volumetric flow of dopant gas to the melt, using a control valve actuated by pressure differences and optionally a carrier gas like argon, with temperature and pressure regulation to maintain consistent dopant supply.
Ensures a uniform and reproducible dopant distribution in single-crystal silicon, reducing resistivity variations and preventing dislocations, allowing for continuous dopant supply and stable crystal growth.
Smart Images
Figure 2025534170000001_ABST
Abstract
Description
[Technical Field]
[0001] The subject of the present invention is a method for producing monocrystalline silicon doped with an n-type dopant by pulling the monocrystalline silicon in a reactor chamber by the CZ method from a melt contained in a crucible. [Background technology]
[0002] Conventional technology / issues Single crystal silicon containing relatively high amounts of n-type dopants is primarily required as a raw material for the production of power semiconductor components.
[0003] Sublimable dopant elements, such as phosphorus, arsenic, or antimony, typically come into contact with the melt in their gaseous state. The challenge here is to ensure a sufficiently high concentration of the dopant in the melt, despite the dopant's high vapor pressure. Furthermore, the introduction of the dopant must be regulated with the utmost precision to prevent the formation of dislocations that would terminate the crystal's monocrystalline growth.
[0004] WO 2021 / 115904 describes a method in which a dopant gas is generated in a sublimation installation and passes together with a carrier gas through a conduit into the reactor chamber, reaching an annular channel at the lower end of the heat shield.
[0005] US Patent Application Publication No. 2010 / 0294999 describes the sublimation of a solid dopant in a heated unit within a reactor chamber and the control of the sublimation rate, as well as the passage of the resulting dopant gas through a conduit with a carrier gas into the melt. Summary of the Invention [Problem to be solved by the invention]
[0006] The object of the present invention is to use simple means to achieve a very uniform dopant distribution over the axial length of the cylindrical portion of the single crystal, regardless of the length of the cylindrical portion. [Means for solving the problem]
[0007] The object is to provide a method for producing monocrystalline silicon doped with n-type dopants by pulling a monocrystalline silicon surrounded by a heat shield having a lower end from a melt contained in a crucible by the CZ method in a reactor chamber, the method comprising: heating the solid dopant in a dopant crucible of a sublimation unit outside the reactor chamber by a crucible heater to a temperature at which a gaseous dopant is formed; and delivering the gaseous dopant in the form of a volumetric flow of dopant gas to the surface of the melt through a conduit having a lower end, the delivering comprising: This is achieved by a method that includes forcing a control valve between the sublimation unit and the reactor chamber to an open state as soon as the pressure difference between the pressure in the sublimation unit and the pressure in the reactor chamber increases to a predetermined value, and controlling the open state of the control valve with the set pressure in the sublimation unit as a command variable and the pressure difference as a controlled variable.
[0008] The present invention allows for a substantially constant flow of dopant gas to be passed through the melt over a relatively long period of time while pulling a single crystal, with solid dopant being recharged as needed to ensure a continuous supply of dopant gas to the melt. This solution is simple and reliable.
[0009] The gaseous dopant is passed into the melt as a volumetric flow of dopant gas, either without or with a carrier gas such as argon. The carrier gas is used to stabilize the flow of the gaseous dopant, especially when the concentration of the gaseous dopant is relatively low. The supply of the carrier gas is preferably controlled by a mass flow controller (MFC).
[0010] The solid dopant may be added to the dopant crucible before and / or during the pulling of the single crystal. To recharge the dopant crucible with solid dopant, the conduit outside the reactor chamber is shut off by completely closing the control valve. For safety reasons, a shutoff valve is preferably activated as well, possibly located between the reactor chamber and the control valve. The shutoff valve ensures reliable access to the sublimation unit, regardless of the tightness of the control valve. The temperature of the dopant crucible, and therefore the temperature in the sublimation unit, is preferably raised just before the solid dopant contained therein is consumed. This ensures that the dopant has completely escaped before the sublimation unit is opened. The escape of the dopant can be monitored via the associated pressure drop in the sublimation unit.
[0011] The sublimation unit comprises a housing that contains and provides access to the dopant crucible, the housing preferably having a double wall construction and capable of being actively cooled by a cooling circuit.
[0012] As well as the crucible heater, other heaters are preferably present, particularly to counteract dopant resublimation. Examples include an internal heater for the housing, an external heater for the control valve and optional shut-off valve, and areas of the conduits located outside the reactor chamber. Additional heaters may be provided to heat the valves that introduce carrier gas and ambient air into the sublimation unit. For red phosphorus as the dopant, the sublimation temperature is 362°C at a pressure of 10,000 Pa, and for arsenic, the sublimation temperature at this pressure is 508°C. In particular, the temperature of the dopant transport conduits must be higher than the corresponding sublimation temperature.
[0013] The pressure is measured in the reactor chamber and in the sublimation unit. The open state of the control valve is determined according to the difference between the pressure in the sublimation unit and the pressure in the reactor chamber and is adjusted via a control valve actuator. The actuator adjusts the stroke of an element that determines the flow through the valve, such as the position of a stopper that can block the flow through the valve. The actuator is preferably electrically operated and forms the control element of the control device. During the supply of gaseous dopant to the surface of the melt, the pressure in the sublimation unit is greater than the pressure in the reactor chamber.
[0014] To ensure that the volumetric flow of dopant gas reaches the melt in a reproducible and controlled manner during the single-crystal silicon pulling procedure, the volumetric flow preferably passes through the control valve only after a stabilization step. The stabilization step involves heating the dopant crucible to a temperature below the temperature at which the solid dopant begins to sublime, with the control valve closed, until the pressure in the sublimation unit rises to a predetermined value. After the stabilization step, the dopant crucible is heated to a temperature above the temperature at which the solid dopant begins to sublime. The rate at which the solid dopant sublimes (evaporation rate) is determined, inter alia, by the temperature to which the dopant crucible is heated. A further pressure increase in the sublimation unit while the temperature of the dopant crucible heater remains constant or increases slightly indicates the onset of sublimation. The control valve opens as soon as the difference between the pressure in the sublimation unit and the pressure in the reactor chamber reaches a predetermined value. The pressure difference is preferably at least 1,000 Pa, more preferably 5,000 to 10,000 Pa. The control valve's opening state is then controlled by a control valve actuator as part of a control process, with the setpoint pressure in the sublimation unit as the command variable and the pressure difference as the control variable. The control preferably takes the form of a PID control. The pressure in the sublimation unit remains substantially constant, except for periodic pressure fluctuations caused by the control, until a pressure drop occurs due to a decrease in the supply of solid dopant in the dopant crucible. When the pressure drops to a desired value, the control valve also closes.
[0015] The dopant crucible can be refilled with solid dopant if it is necessary to replace dopant that has escaped from the melt. This can occur when the crystal is remelted after dislocation generation to begin a new single crystal pulling run, or when the dopant stock in the sublimation unit is depleted due to the length of the cylindrical portion of the single crystal being grown. In this case, with the control valve and any isolation valves closed, the sublimation unit is purged and cooled, and the required amount of additional solid dopant is placed in the dopant crucible.
[0016] The supply of the volumetric flow of dopant gas through the conduit to the surface of the melt is preferably initiated before the cylindrical portion of the single crystal is pulled, for example, immediately after the seed crystal is brought into contact with the melt or before or during the pulling of the conical portion of the single crystal. An alternative possibility is to start the method only at the beginning of the pulling of the cylindrical portion. Regardless of this, the melt may be enriched with dopant even before the supply of the volumetric flow of dopant gas, for example, by immersing a doping bell with solid dopant in the silicon melt.
[0017] The lower end of the conduit, from which the volumetric flow of dopant gas emerges, is preferably positioned above the melt at the same height as the lower end of the heat shield surrounding the growing single crystal. Furthermore, the lower end of the conduit is preferably located between the lower end of the heat shield and the wall of the crucible containing the melt. It is also preferred that the conduit within the reactor chamber be positioned so that it passes through the wall of the heat shield. In the region of the reactor chamber wall where the conduit can be cooled by water, the conduit is preferably made of a metal, such as stainless steel, and is exposed to temperatures of up to 600°C. Around the growing single crystal and the relatively hot graphite section, temperature-stable materials for the high-temperature zone, such as graphite, carbon fiber-reinforced composite (CFC) materials, and / or quartz, should be used. It is particularly preferred to divide the conduit within the reactor chamber into an upper section made of stainless steel, a middle section made of CFC, and a lower section made of quartz clad with CFC at its lower end. The cladding protects the quartz tube from overheating.
[0018] The invention will be further explained below with reference to the drawings. [Brief explanation of the drawings]
[0019] [Figure 1] 1 is a vertical cross-sectional view of a reactor chamber and sublimation unit suitable for practicing the present invention. [Figure 2] FIG. 2 is an enlarged detail view from FIG. [Figure 3] FIG. 1 shows the profile of electrical resistivity R of single crystal silicon as a function of axial position P of the cylindrical portion of the single crystal. DETAILED DESCRIPTION OF THE INVENTION
[0020] DETAILED DESCRIPTION OF EXEMPLARY EMBODIMENTS IN ACCORDANCE WITH THE INVENTION FIG. 1 shows a vertical cross section of a reactor chamber 1 and a sublimation unit 2 suitable for implementing the present invention. The following description does not refer to some of the features shown, which do not contribute to the description of the present invention. A crucible 3 containing a silicon melt 4 is placed within the reactor chamber 1. The silicon melt 4 is doped with a gaseous n-type dopant from the melt during the pulling of a single crystal 5. The gaseous n-type dopant is passed through a conduit 6 to the surface of the melt 4. The melt 4 is kept liquid by a heating device 7 surrounding the crucible 3. The single crystal 5 is surrounded by a heat shield 10 that blocks the thermal radiation of the heating device 7. The crucible 3 is raised and rotated while the single crystal 5 is rotated and pulled from the melt by a pulling device 8. It is possible, but not absolutely necessary, to apply a magnetic field, such as a horizontal magnetic field, to the melt 4 by one or more magnetic field coils 9 arranged around the reactor chamber 1. During the pulling of the single crystal 5, the reactor chamber 1 is purged, for example, by a flow of argon passing through it.
[0021] The gaseous dopant is generated in a sublimation unit 2 located outside the reactor chamber, from where it is guided in the form of a volumetric flow of dopant gas through a conduit 6 to the surface of the melt 4. The volumetric flow may include a portion of a carrier gas (e.g., argon) that is passed to the sublimation unit via a mass flow controller 14. To generate the gaseous dopant, a solid dopant, e.g., red phosphorus, is introduced into a dopant crucible 11, which is heated by a crucible heater 12 in the sublimation unit 2. The sublimation unit 2 may possess a weighing cell 13 that provides information about the current weight of the solid dopant in the dopant crucible 11. On its path through the conduit 6, after leaving the sublimation unit 2 and before entering the reactor chamber 1, the volumetric flow of dopant gas passes through a control valve 16 and, optionally, a shut-off valve 17.
[0022] The pressure and temperature in the reactor chamber 1 and the sublimation unit 2 are transmitted by sensors to a controller 15, which is used to control the supply of dopant gas to the surface of the melt 4. In the illustrated embodiment, the controller 15 is also used to guide the pulling operations, e.g., the rotation and removal of the single crystal 5 and the rotation and lifting of the crucible 3. To begin the supply of gaseous dopant to the surface of the melt 4, the dopant crucible 11 is heated by a crucible heater 12. During this time, the controller 15 keeps the control valve 16 and the shut-off valve 17 (if present) closed. The solid dopant in the dopant crucible 11 is first preheated to a temperature below the sublimation temperature until the pressure in the sublimation unit 2 reaches a predetermined threshold. Subsequently, the temperature of the crucible heater 12 is increased to a target temperature higher than the sublimation temperature, resulting in a further increase in the pressure in the sublimation unit 2. The controller 15 forces the control valve 16 to the open state and, if necessary, opens the shut-off valve 17 as soon as the difference between the pressure in the sublimation unit 2 and the pressure in the reactor chamber 1 reaches a predetermined value. The open state of the control valve 16 is then controlled, with the setpoint pressure in the sublimation unit 2 serving as the control command variable. As the stock of solid dopant in the dopant crucible decreases, the pressure in the sublimation unit 2 decreases. Once the pressure drops to the forced value, the control valve 16 and, if present, the shut-off valve 17 are closed. If necessary, the dopant crucible 11 can then be filled with more solid dopant, even during the pulling process of the single crystal 5, and a new cycle of supplying gaseous dopant to the surface of the melt 4 can begin.
[0023] 1, the lower end of the conduit 6 has the same distance from the surface of the melt 4 as the lower end of the heat shield 10. Furthermore, in this embodiment, the profile of the conduit 6 in the reactor chamber 1 is designed so that the conduit 6 passes through the heat shield 10 and the lower end of the conduit 6 is located between the wall of the crucible 3 and the lower end of the heat shield 10.
[0024] According to this preferred configuration and representation in Figure 2, the conduit 6 within the reactor chamber 1 is divided into three sections: an upper section 18 made of stainless steel, a middle section 19 made of CFC (carbon fiber composite), and a lower section 20 made of quartz surrounded at its lower end by a CFC jacket.
[0025] The effectiveness of the invention was tested, by way of example, using the production of monocrystalline silicon having a diameter of 200 mm in a facility substantially including the features depicted in Figure 1. The sublimation unit was manufactured by the French manufacturer Riber.
[0026] First, a single crystal was pulled up to the point of the upper cylindrical section, which had a basic phosphorus doping that had been added to the melt before pulling by a doping bell. Using the method of the present invention, it was shown that the electrical resistivity of the single crystal could be further reduced and kept substantially constant between the lower LSL and upper USL until the end of the pulling procedure. Figure 3 presents the results in the form of resistivity measurements. The resistivity profile is shown from the beginning of the cylindrical section at position P=0 to the end of the cylindrical section at position P=100. The supply of gaseous dopant according to the present invention, in the form of a volumetric flow of dopant gas, was carried out during the pulling of the cylindrical section of the single crystal between the positions indicated as the beginning and the end. By applying the method of the present invention, it was possible to reduce the resistivity in the single crystal to less than 1 mΩ cm. [Explanation of symbols]
[0027] List of reference symbols used 1. Reactor chamber 2 sublimation units 3 Crucible 4. Melt 5. Single crystal 6 Conduit 7 Heating device 8. Lifting device 9 magnetic field coil 10 Heat Shield 11 Dopant crucible 12 Crucible heater 13 Weighing cell 14 Mass flow controller 15 Controller 16 Control valve 17 Shut-off valve 18 Upper part 19 Middle part 20 Lower part 21 CFC Jacket
Claims
1. 1. A method for producing single crystal silicon doped with an n-type dopant by pulling a single crystal surrounded by a heat shield having a lower end from a melt contained in a crucible by the CZ method in a reactor chamber, the method comprising: heating a solid dopant in a dopant crucible of a sublimation unit outside the reactor chamber by a crucible heater to a temperature at which a gaseous dopant is formed; and supplying the gaseous dopant in the form of a volumetric flow of dopant gas through a conduit having a lower end to a surface of the melt, said supplying comprising: a control valve between the sublimation unit and the reactor chamber being forced to an open state as soon as a pressure difference between the pressure in the sublimation unit and the pressure in the reactor chamber increases to a predetermined value, and controlling the open state of the control valve using a set pressure in the sublimation unit as a command variable and the pressure difference as a controlled variable.
2. The method of claim 1 , wherein the control takes the form of a PID control.
3. The method of claim 1 or 2, wherein the lower end of the conduit has the same distance from the surface of the melt as the lower end of the heat shield.
4. The method of any one of claims 1 to 3, wherein the conduit passes through the heat shield, and the lower end of the conduit is located between a wall of the crucible and the lower end of the heat shield.
5. The method according to any one of claims 1 to 4, wherein the conduit is divided into an upper section made of stainless steel, a middle section made of CFC, and a lower section made of quartz clad with CFC.
6. The method according to any one of claims 1 to 5, wherein the conduit is closed between the sublimation unit and the reactor chamber by an isolation valve before the dopant crucible is reloaded with more solid dopant.
7. The method of any one of claims 1 to 6, wherein the temperature of the dopant crucible is increased just before the solid dopant contained therein is depleted.