Method for producing a silicon ingot and silicon ingot

Abstract

Method of Czochralski, CZ -growth of a silicon ingot (112) for IGBTs, wherein the method comprises: Melting a mixture of a silicon material and an n-type dopant material in a crucible (105), Extraction of the silicon ingot (112) from the molten mixture during an extraction time by CZ growth, whereby the silicon ingot (112) is doped with the n-type dopant material of the molten mixture, and Doping the silicon ingot (112) with an additional n-type dopant material during at least one sub-period of the extraction time, wherein the doping of the silicon ingot (112) with the additional n-type dopant material is in addition to the doping with the n-type dopant material of the molten mixture, and the doping of the silicon ingot (112) with the additional n-type dopant material begins after the growth of a first ingot part and ends before the growth of a further ingot part, wherein a net n-type doping in the silicon ingot (112) is between 1 × 10 13 cm -3 and 1 × 10 15 cm -3 is being discontinued.

Description

BACKGROUND

[0001] Silicon wafers grown using the Czochralski (CZ) process, such as the standard CZ process, the magnetic CZ (MCZ) process, or the continuous CZ (CCZ) process, serve as a base material for manufacturing a wide variety of semiconductor devices and integrated circuits, including power semiconductor devices and solar cells. In the Czochralski process, silicon is heated in a crucible to its melting point of approximately 1416°C to create a molten silicon solution. A small silicon seed crystal is brought into contact with the melt. Molten silicon solidifies onto the seed crystal. By slowly withdrawing the seed crystal from the melt, a crystalline silicon ingot, or bar, with a diameter of one or more hundred millimeters and a length of one meter or more, is grown.In the MCZ process, an external magnetic field is additionally applied to reduce oxygen contamination levels.

[0002] The growth of silicon with a defined doping concentration using the Czochralski method is complicated by segregation effects. The segregation coefficient of a dopant material characterizes the relationship between the concentration of the dopant material in the growing crystal and that of the melt. Typically, dopant materials have segregation coefficients lower than one, meaning that the solubility of the dopant material in the melt is greater than in the solid state. This typically leads to an increase in the doping concentration in the ingot with increasing distance from the seed crystal.

[0003] Since, in Czochralski-grown silicon ingots, the tolerance range of a doping concentration or resistivity along the axial direction between opposite ends of the silicon ingot can be smaller than the variability of a doping concentration or resistivity caused by segregation effects during CZ growth, depending on the application of the grown silicon, different parts of the silicon ingots can be used as base materials. These parts can have different target doping concentrations with overlapping, adjacent, or spaced-apart tolerance ranges for a doping concentration or resistivity. Such partitioning of the silicon ingot is also known as order matching.

[0004] A method for producing photovoltaic-suitable crystalline silicon by adding dopants is known from US 2011 / 0030793 A1. Methods for producing a silicon ingot using the Czochralski method are known from WO 2012 / 142463 A2 and the subsequently published DE 10 2015 100393 A1 and DE 10 2014 107590 B3. An exemplary device for pulling single crystals is known from DE 38 06918 A1.

[0005] It is desirable to provide a silicon ingot and a method for producing a silicon ingot grown by the Czochralski process, which allows for an improved yield of silicon ingot parts having doping concentrations or specific resistances that are within acceptable tolerance ranges.

[0006] It is therefore an object of the present invention to provide a method for producing a silicon ingot and a silicon ingot, each of which satisfies the above requirements.

[0007] This problem is solved according to the invention by a method with the features of claim 1 or by a silicon ingot with the features of claim 22. Advantageous embodiments of the invention are set forth in the dependent claims. SUMMARY

[0008] According to one embodiment, a method for Czochralski growth of a silicon ingot is disclosed. The method comprises melting a mixture of a silicon material and an n-type dopant material in a crucible. The silicon ingot is extracted or removed from the molten silicon during an extraction or removal period. The method further comprises doping the silicon ingot with additional n-type dopant material during at least one sub-period of the extraction period.

[0009] According to one embodiment, an n-doped silicon ingot is disclosed. The resistivity ρ of such a doped silicon ingot has at least one inflection point along an axis between opposite ends of the silicon ingot, where a concavity or curvature of the resistivity ρ changes along the axis.

[0010] The expert will recognize additional features and advantages after reading the following detailed description and examining the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The accompanying drawings are enclosed to provide a further understanding of the invention and are incorporated into the disclosure and form part thereof. The drawings illustrate exemplary embodiments of the present invention and, together with the description, serve to explain the principles of the invention. Other exemplary embodiments and intended advantages are immediately appreciated, as they are better understood with reference to the following detailed description. Fig. Figure 1 is a schematic flowchart to illustrate a process for manufacturing an n-type silicon ingot. Fig.2 is in graph, which has a specific resistance ρ as a function of an axial direction a of a by the method of Fig. represents 1 grown silicon ingots. Fig. Figure 3 is a schematic sectional view of a CZ growth system for executing the in Fig. 1. Procedure shown. Fig. Figure 4 is a schematic cross-sectional view of a crucible to illustrate a method for doping the crucible with dopant material. Fig. Figure 5 is a schematic sectional view of a part of a CZ growth system to illustrate a process of adding dopants to a silicon melt in the crucible. Fig.Figure 6 is a graph illustrating a simulated concentration of uncompensated phosphorus along an axial position of a CZ-grown silicon ingot with respect to different ratios of boron and phosphorus added to the silicon melt. Fig. Figure 7 is a graph illustrating a simulated resistivity along an axial position of a CZ-grown silicon ingot with respect to different ratios of boron and phosphorus added to the silicon melt. DETAILED DESCRIPTION

[0012] The following detailed description refers to the accompanying drawings, which form part of the disclosure and show specific embodiments for illustrative purposes in which the invention can be implemented. It is understood that other embodiments can be used and structural or logical modifications can be made without departing from the scope of the present invention. For example, features illustrated or described for one embodiment can be used in or in conjunction with other embodiments to arrive at yet another embodiment. The present invention is intended to encompass such modifications and changes. The examples are described using specific language, which should not be interpreted as limiting the scope of the appended claims.The drawings are not to scale and are for illustrative purposes only. For clarity, the same elements are identified by corresponding reference symbols in the various drawings, unless otherwise stated.

[0013] The terms "have," "contain," "comprise," "exhibit," and similar terms are open-ended, indicating the presence of the identified structures, elements, or features, but not excluding the presence of additional elements or features. Indefinite and definite articles should indicate both the plural and singular unless the context clearly indicates otherwise.

[0014] The term "electrically connected" describes a permanent, low-resistance connection between electrically connected elements, for example, a direct contact between the elements in question or a low-resistance connection via a metal and / or a highly doped semiconductor. The term "electrically coupled" implies that one or more intermediate elements, adapted for signal transmission, may be present between the electrically coupled elements, for example, elements that provide a low-resistance connection in a first state and a high-resistance electrical decoupling in a second state.

[0015] The figures illustrate relative doping concentrations by indicating "-" or "+" next to the doping type "n" or "p". For example, "n" means -“a doping concentration that is lower than the doping concentration of an “n” doping range, while an “n + A doping area with a relative concentration has a higher doping concentration than an n-type doping area. Doping areas with the same relative concentration doping do not necessarily have the same absolute concentration. For example, two different n-type doping areas can have the same or different absolute concentrations of doping.

[0016] Fig. 1 refers to a method for producing a silicon ingot or silicon bar or silicon block.

[0017] One process feature S100 of the process includes melting a mixture of a silicon material and an n-type dopant material in a crucible.

[0018] One process feature S110 of the process includes extracting or removing the silicon ingot from the molten silicon over an extraction period.

[0019] A process feature S120 of the process includes doping the silicon ingot with additional n-type dopant material during at least one sub-period of the extraction time.

[0020] By doping the silicon ingot with the additional n-type dopant material during at least one sub-period of the extraction time, the production of at least two ingot parts having different target specifications of a specific resistance ρ, also known as order matching, can be improved, as shown in the schematic graph of Fig. Figure 2 illustrates this. The schematic graph of Fig.Figure 2 shows the resistivity ρ of the silicon ingot as a function of an axial direction a. Growth of the silicon ingot began at the axial position 0%. Due to segregation effects during growth of the n-type silicon ingot, the resistivity slope increases negatively towards one end of the silicon ingot. By intermittently doping with the additional n-type dopant material after the growth of a first silicon ingot part P1, which meets a first target resistivity specification, the axial distance da between the first silicon ingot part P1 and a second silicon ingot part P2, which meets a different target resistivity specification, can be reduced compared to a second reference ingot part P2REF, which grew without doping with the additional n-type dopant material.Temporary doping with the additional n-type dopant material leads to an increase in the negative slope of the resistivity along an axial direction of the silicon ingot. By performing the doping with the additional n-type dopant material in a resistivity range Δρ1 between the first and second target specifications, the axial distance da between the first and second silicon ingot parts P1, P2, which meet the first and second target specifications, can be reduced compared to the axial distance daref between the first ingot part P1 and the second reference ingot part P2REF. In other words, the second silicon ingot part P2 is shifted towards one of the origins of the silicon ingot.Since the negative inclination of the resistivity ρ along the axial direction a to the origin of the silicon ingot decreases due to segregation effects, the extent of the second silicon ingot part P2, which meets the second target specification, can be increased compared to the second reference ingot part P2REF in order to improve the yield.

[0021] According to one embodiment, the additional n-type dopant material is phosphorus.

[0022] According to another embodiment, the silicon ingot is doped with the additional n-type dopant material by a vapor-phase doping technique. One embodiment includes controlling an inlet of a dopant precursor gas into a reaction chamber containing the silicon ingot. Phosphine (PH3) and arsine (AsH3) are examples of precursor gases for n-type doping of silicon.

[0023] According to another embodiment, doping the silicon ingot with the additional n-type dopant material comprises melting an n-type dopant source material in the crucible. The silicon ingot can be doped with the additional n-type dopant material by setting a depth of the n-type dopant source material in the molten silicon, wherein the n-type dopant source material comprises the additional n-type dopant material. Setting the depth of the n-type dopant source material immersed in the molten silicon in the crucible may include measuring a weight of the n-type dopant source material.

[0024] According to another embodiment, the n-type dopant source material is in the form of one or more rods. These rods can be immersed in the silicon melt by a puller mechanism. The puller mechanism holds the n-type dopant source material, immerses it in the silicon melt, and also withdraws it from the silicon melt. A control mechanism is provided to control the puller mechanism. This control mechanism can, for example, control the puller mechanism by wired or wireless control signal transmission.

[0025] According to one embodiment, the n-type dopant source material is made of quartz or silicon carbide doped with the additional n-type dopant material. The concentration profile of the additional n-type dopant material at a certain depth within the n-type dopant source material can have a peak value below a surface of the n-type dopant source material. The introduction of the n-type dopant material into the n-type dopant source material can be carried out by at least one of the following processes: in-situ doping, plasma deposition or deposition through a surface of the n-type dopant source material, ion implantation through the surface of the n-type dopant source material, and diffusion through the surface of the n-type dopant source material.The addition of n-type dopant source material that has a peak value below the surface of the n-type dopant source material leads to a delay in the addition of additional n-type dopants to the melt, depending on the depth profile of a doping and the melting rate of the n-type dopant source material.

[0026] According to one embodiment, the degree of doping of the silicon ingot with the additional n-type dopant material is varied or changed between no doping and maximum doping during the extraction time. Doping of the silicon ingot with the additional n-type dopant material can be suppressed by stopping additional n-type dopants entering the molten silicon, for example by pulling out the n-type dopant source material, e.g., a rod, from the molten silicon via the puller mechanism, or by stopping the inlet of the dopant precursor gas into the reaction chamber containing the silicon ingot.

[0027] According to the invention, the net n-type doping in the silicon ingot is between 1 × 10 13 cm -3 and 1 × 10 15 cm -3discontinued. Raw materials, for example wafers for power semiconductors such as IGBTs, diodes, insulated-gate field-effect transistors (IG-FETs) and thyristors, can have a net n-type doping range between, for example, 1 × 10 13 cm -3 and 1 × 10 15 cm -3 have.

[0028] According to another embodiment, the silicon ingot is doped with p-type dopant material by adding the p-type dopant material to the silicon melt using a p-type dopant source material and vapor-phase doping. In one embodiment, the p-type dopant material is a boron-aluminum-gallium compound. Doping with the p-type dopant material partially compensates for the n-type doping. Due to the different segregation of the n- and p-type dopants, a further decrease in the negative slope of the resistivity along the axial direction a can be achieved. If the p-type dopants are added during a growth period of the first and second ingot parts P1, P2, the decrease in the negative slope coincides with these periods. This allows for an increased expansion of the first and second ingot parts P1, P2 along the axial direction, leading to an improved yield.Since the method described above allows a displacement of silicon ingot parts of a specified resistivity along the axial direction, other properties of silicon ingot parts that change along the axial direction, such as oxygen content, can be adjusted by appropriately displacing the target silicon ingot parts along the axial direction.

[0029] According to one embodiment, the silicon ingot is doped with the p-type dopant material by a vapor phase doping technique.

[0030] According to another embodiment, doping the silicon ingot with the p-type dopant material comprises melting p-type dopant source material in the crucible. The silicon ingot can be doped with the p-type dopant material by setting a depth of the p-type dopant source material in the molten silicon, wherein the p-type dopant source material comprises the p-type dopant material. Setting the depth of the p-type dopant source material immersed in the molten silicon in the crucible can include measuring a weight of the p-type dopant source material. The p-type dopant source material can be in the shape of one or more rods. The p-type dopant source material can be made of quartz or silicon carbide doped with the additional n-type dopant material.The concentration profile of the p-type dopant material at a certain depth within the p-type dopant source material can have a peak below a surface of the p-type dopant source material. The p-type dopant material can be introduced into the p-type dopant source material by at least one of the following processes: in-situ doping, plasma deposition through a surface of the p-type dopant source material, ion implantation through the surface of the p-type dopant source material, and diffusion through the surface of the p-type dopant source material. The degree of doping of the silicon ingot with the additional n-type dopant material during the extraction time can be varied from no doping to maximum doping.

[0031] Fig. Figure 3 is a simplified schematic sectional view of a CZ growth system 100 for executing the in Fig. 1. Procedure shown.

[0032] The CZ growth system 100 comprises a crucible 105, for example a quartz crucible on a crucible support 106, e.g. a graphite susceptor. A heating device 107, e.g. a high-frequency (HF) coil, surrounds the crucible. The heating device 107 can be arranged on lateral sides and / or on a bottom side of the crucible 105. The crucible 105 can be rotatable by means of a bearing shaft 108.

[0033] The mixture of silicon material, e.g., a non-crystalline raw material such as polysilicon, and an n-type dopant material, such as phosphorus (P), antimony (Sb), arsenic (As), or any combination thereof, is melted in the crucible by heating via the heating device 107. The n-type dopant material can already constitute the initial doping of the silicon material to be melted, be a part of it, and / or be added as a solid, liquid, or gaseous dopant source material. According to one embodiment, the solid dopant source material is a dopant source particle, such as a dopant source pill or tablet. The dopant source material can have a predetermined shape, such as a disk shape, a spherical shape, or a cubic shape. For example, the shape of the dopant source material can be adapted to a supply device 109, such as a dispenser.The dispenser is adapted to deliver the dopant source material to a silicon melt 110 in the crucible 105.

[0034] According to one embodiment, the dopant source material can comprise a support material or a binder material in addition to the dopant material. For example, the dopant source material can be quartz or silicon carbide (SiC) doped with the dopant material. According to another embodiment, the dopant source material can be highly doped silicon material, such as highly doped polysilicon material, which is doped to a greater extent than the silicon raw material. According to yet another embodiment, the dopant source material can be boron nitride and / or boron carbide.

[0035] A silicon ingot 112 is drawn from the crucible 105 containing the silicon melt 110 by immersing a seed crystal 114 into the silicon melt 110, which is then slowly withdrawn at a surface temperature of the melt just above the melting point of silicon. The seed crystal 114 is a single-crystal silicon seed mounted on a seed carrier 115, which is rotated by a drawing shaft 116. A drawing rate, typically in the range of a few mm / min, and a temperature profile influence the diameter of the CZ-grown silicon ingot 112.

[0036] If the silicon ingot 112 with the CZ growth system 100 according to the in Fig. In the process illustrated in 1, the additional n-type dopant material is added to the silicon melt during at least one sub-period of the extraction time.

[0037] According to another embodiment, the additional n-type dopant material is temporarily added to the molten silicon from a doped quartz material, such as a phosphorus-doped quartz material, which is supplied to the silicon melt 110 by the supply device 109.

[0038] According to another embodiment, the additional n-type dopant material is added to the silicon melt 100 from a doped crucible. The crucible doped with additional n-type dopant material can be formed by implanting the additional n-type dopant material, for example phosphorus, into the crucible (see the schematic sectional view of Fig. 4) The additional n-type dopant material can be inserted into crucible 105 through one or more inclined implantations, see reference numeral I2 2 and I3 2 , and / or by non-angled implantation, see reference I1 2 in Fig.4. A distribution of an inclination angle or inclination angles can be used to adjust the amount of additional n-type dopant material supplied to the silicon melt 110 by dissolving material from the crucible 105 in the silicon melt 110, for example, at a rate in the range of approximately 10 µm / hour if the crucible is made of quartz. The additional n-type dopant material can be implanted in the crucible at different energies and / or doses. Applying a thermal budget to the crucible 105 by heating can allow the establishment of a retrograde profile in the additional n-type dopant material within the crucible 105. Multiple implantations at different energies and / or doses further allow the establishment of a profile of the additional n-type dopant material at a depth within the crucible 105.Thus, the addition rate of the additional n-type dopant material to the silicon melt 110 can be set; that is, by selecting implantation parameters, the addition rate of the additional n-type dopant material can be changed and controlled in a well-defined manner. For example, the profile of the additional n-type dopant material in the crucible 105 can be a retrograde profile. As an alternative or in addition to implanting the additional n-type dopant material into the crucible 105, the additional n-type dopant material can also be introduced into the crucible 105 by another process, e.g., by diffusion from a diffusion source, such as a solid diffusion source of the additional n-type dopant material.As a further alternative or in addition to the above processes of introducing the additional n-type dopant material into the crucible 105, the additional n-type dopant material can also be introduced into the crucible 105 in situ, i.e. during the formation of the crucible 105.

[0039] According to yet another embodiment, the additional n-type dopant material can be introduced into the silicon melt 110 from the gas phase, for example, by supplying phosphine (PH3) as a precursor gas for n-type doping of silicon via the supply device 109. According to one embodiment, the supply of boron in the gas phase can occur via a supply of inert gas into the CZ growth system 100. According to another embodiment, the additional n-type dopant material can be supplied in the gas phase via one or more tubes, for example, a quartz tube extending into the silicon melt 110. According to yet another embodiment, the additional n-type dopant material can be supplied in the gas phase via one or more tubes that terminate at a short distance from the silicon melt 110.The pipes can include one or more openings at an outlet, for example in the form of a shower head.

[0040] According to another embodiment, a lining or lining layer can be formed on the crucible 105 to control the diffusion of the additional n-type dopant material from the crucible 105 into the silicon melt 110. For example, the lining layer can be made of quartz and / or silicon carbide. According to one embodiment, the lining layer can dissolve in the silicon melt 100 before the additional n-type dopant material contained in the crucible dissolves into the silicon melt 110 and serves as a dopant during the growth of the silicon ingot 112. This allows for the timing of when the additional n-type dopant material is available in the silicon melt as a dopant to be introduced into the silicon ingot 112.The lining layer can also delay the introduction of the additional n-type dopant material into the silicon melt 110 for a period of time required for diffusion of the additional n-type dopant material from the crucible 105 through the lining layer and into the silicon melt 110.

[0041] According to another embodiment, the method for producing the silicon ingot 112 further comprises changing the addition rate of the additional n-type dopant material to the silicon melt 110. In one embodiment, changing the addition rate of the additional n-type dopant material to the silicon melt 110 comprises changing at least one parameter from the dimensions, geometry, and delivery rate of particles, including the additional n-type dopant material. For example, the rate can be increased by increasing the diameter of the particles doped with the dopant material. As an additional or alternative measure, the addition rate of the additional n-type dopant material to the silicon melt 110 can be increased by increasing the feed rate of the dopant source material into the silicon melt 110 by the supply device 109.

[0042] According to another embodiment, which is shown in the schematic sectional view of Fig. As illustrated in Figure 5, changing the addition rate of the additional n-type dopant material to the silicon melt 110 involves changing the depth d of a dopant source material 125 immersed in the silicon melt 110.

[0043] According to another embodiment, changing the addition rate of the additional n-type dopant material to the silicon melt 110 involves changing the temperature of the dopant source material 125. For example, increasing the temperature of the dopant source material, e.g., by heating it, can increase the amount of additional n-type dopant material introduced into the silicon melt 110 from the dopant source material 125. The dopant source material 125 is doped with the additional n-type dopant material. According to one embodiment, doping of the dopant source material is carried out by a process consisting of in-situ doping, plasma deposition, or...The process involves the deposition of the dopant source material 125 through a surface 126, ion implantation through the surface 126 of the dopant source material 125, and diffusion through the surface 126 of the dopant source material 125. The dopant source material 125 can be shaped, for example, as a rod, cylinder, cone, or pyramid. It can also be made from a multitude of separate dopant source pieces having one shape or a combination of different shapes. The depth d of a portion of the dopant source material 125 immersed in the silicon melt 110 can be changed by a puller mechanism 127. The puller mechanism 127 holds the dopant source material 125, immerses the dopant source material 125 in the silicon melt 110 and also pulls the dopant source material 125 out of the silicon melt 110.A control mechanism 128 is designed to control the puller mechanism 127. The control mechanism 128 can control the puller mechanism 127, for example, by wired or wireless control signal transmission.

[0044] According to another embodiment, changing the addition rate of the additional n-type dopant material to the silicon melt 110 includes changing a flux or partial pressure of a precursor gas, e.g. phosphine (PH3), when the silicon melt 110 is doped with boron from the gas phase.

[0045] The above-described method for manufacturing the silicon ingot 112 leads to improved order matching or to improved adaptation of the sequence of at least two ingot parts that have different target specifications of a specific resistance ρ.

[0046] Partial counter-doping with p-type dopant material can be carried out using similar techniques to those described above for additional n-type doping. Doping with p-type dopant material leads to partial compensation of the n-type doping. Due to the different segregation of the n- and p-type dopant materials, a further decrease in the negative slope of resistivity along the axial direction can be achieved. If p-type dopant materials are added during a growth period of ingot portions that fall within a specific range of a target resistivity, the extension of these ingot portions along the axial direction can be increased, leading to improved yield. Further details on a decrease in the negative slope of resistivity along the axial direction are given below.

[0047] An axial profile of doping caused by segregation of dopant material during CZ growth can be approximated by equation (1) below: c(p)=k0c0(1−p)k0−1+F0k01−k0[(1−p)k0−1−1]

[0048] The first term in equation (1) refers to doping added to the melt before the silicon ingot is extracted or withdrawn from the melt. According to the above embodiments, a basic n-type dopant material can be described by the first term of equation (1). The second term refers to adding dopant material to the melt at a constant rate during CZ growth. According to the above embodiments, adding boron or any other additional dopant material can be described by the second term of equation (2).

[0049] In the equation (1) above, c(p) denotes a concentration of dopant material in the silicon ingot (atoms / cm²). 3 ), p denotes a portion of the initial melt during a CZ growth that has crystallized and corresponds to an axial position between 0% and 100% of the fully grown silicon ingot, k0 denotes a segregation coefficient of the dopant material, e.g. approximately 0.8 for boron (B) in silicon and approximately 0.35 for phosphorus (P) in silicon, c0 denotes an initial concentration of the dopant material in the melt (atoms / cm³). 3 ) and F0 denotes the total amount of dopant material added to the melt at a constant rate (with respect to the drawing rate), divided by the initial volume of the melt (atoms / cm³). 3 ).

[0050] Fig.Figure 6 illustrates calculated concentrations of uncompensated phosphorus (P), that is, net n-doping, as a function of axial position between opposite ends of a silicon ingot. The curves shown refer to different ratios of boron (B) and phosphorus (P), that is, F OB / c OP corresponding to the ratio of the total amount of boron added to the silicon melt at a constant rate (relative to the drawing rate) divided by the initial volume of the melt (F OB in atoms / cm² 3 ) and an initial concentration of phosphorus in the melt (c OP in atoms / cm² 3 ).

[0051] The curves shown refer to values ​​of F OB / c OP of 0%, 10%, 20%, 30%, 40%, 50%.

[0052] Fig.Figure 7 illustrates calculated specific resistance curves as a function of axial position between opposite ends of a silicon ingot. Similar to those in Fig. The parameter curves shown in 6 refer to those in Fig. The 7 curves shown represent different ratios of boron (B) and phosphorus (P), that is, F OB / c OP corresponding to the ratio of the total amount of boron added to the silicon melt at a constant rate (relative to the drawing rate) divided by the initial volume of the melt (F OB in atoms / cm² 3 ) and an initial concentration of phosphorus in the melt (c OP in atoms / cm² 3 ) .

[0053] Similar to those in Fig. The parameter curves shown in 6 refer to those in Fig. The 7 curves shown have values ​​of F OB / c OPof 0%, 10%, 20%, 30%, 40%, 50%. By adding boron to the melt during CZ growth, and thus adding a compensating dopant to the melt during CZ growth, this allows, based on Fig. The methods described in sections 1 to 5 reduce the negative slope of the resistivity along the axial direction between opposite ends of the silicon ingot. Since the use of silicon ingots for feeding wafers to manufacture a semiconductor device may require small tolerances with respect to resistivity, for example, for the production of insulated-gate bipolar transistors (IGBTs), the methods described in sections 1 to 5 allow for a reduction of the negative slope of the resistivity along the axial direction between opposite ends of the silicon ingot. Fig. 6 and Fig.The methods described in section 7 result in a yield improvement. An axial length PREF of a silicon ingot part without p-type counter-doping, which falls within a specific resistance range between ρ1 and ρ2, is smaller than an axial length P of a silicon ingot part with 10% p-type counter-doping, which falls within the specific resistance range between ρ1 and ρ2.

[0054] Based on the following Fig. Table 1 shows a maximum part of the ingot along the axial direction, which has a specific fluctuation of a specific resistance and a specific ratio of boron (B) and phosphorus (P), that is, F, in the processes illustrated and described in sections 1 to 7. OB / c OP corresponding to the ratio of the total amount of boron added to the silicon melt at a constant rate (relative to the drawing rate) divided by the initial volume of the melt (F OB in atoms / cm² 3) and an initial concentration of phosphorus in the melt (c OP in atoms / cm² 3 Table 1 refers to values ​​of F OB / c OP of 0%, 10%, 20%, 30%, 40%, 50% and axial fluctuations of resistivity of + / - 5%, + / - 10%, + / - 15%, + / - 20%, + / - 30%, + / - 50%. By adding boron to the melt during CZ growth, and thus adding a compensating dopant to the melt during CZ growth, this allows, based on Fig. The methods described in sections 1 to 6 improve yield by increasing the maximum portion of the ingot along the axial direction that exhibits a specific fluctuation in resistivity. For example, the axial portion of the ingot exhibiting a specific resistivity fluctuation of ±10% can be increased from 26% (no compensation doping) to 78% (compensation doping F). OB / c OP will be increased by 40%). Table 1: maximum ingot length with axial fluctuation of a specific resistance of Boron compensation flux / initial doping with phosphorus + / -5 % + / - 10% + / - 15% + / - 20% + / - 30% + / - 50% No compensation 14 % 26 % 36 % 46 % 60 % 80 % 20 % 32 % 48 % 58 % 66 % 76 % 88 % 30 % 56 % 66 % 74 % 78 % 84 % 92 % 35 % 66 % 74 % 78 % 82 % 86 % 92 % 40 % 38 % 78 % 82 % 84 % 88 % 92 % 45 % 22 % 44 % 84 % 86 % 88 % 94 %

[0055] According to the based on Fig. 6 and Fig. In the 7 illustrated methods, boron becomes constant (relative to the drawing rate) to the silicon melt (described by the term F). OB in atoms / cm² 3 ) is added, and phosphorus is added as an initial concentration to the melt (described by the term c). OP in atoms / cm² 3 ) added. According to other embodiments, boron can be added to the melt at a changing rate. Apart from or in addition to phosphorus, the additional n-type dopant material, such as phosphorus, antimony, or arsenic, can be added in ingot growth intervals between ingot portions that fall within a specific target resistance, as can be seen, for example, from Fig. 1 and Fig. 2 is described.

[0056] In addition to adding boron to the melt during CZ growth, a portion of the total boron can also be added to the melt before CZ growth, as described by a term c. OP can be described in equation (1). Similarly, in addition to adding phosphorus or the other n-type dopant material as an initial concentration to the melt, some of the phosphorus or the additional n-type dopant material can also be added to the melt during a CZ growth, which can be described by a term F. OP can be described in equation (1) if the phosphorus or other n-type dopant material is added at a constant rate relative to the extraction rate.

[0057] According to one embodiment of a silicon ingot, a resistivity ρ along an axis between opposite ends of the silicon ingot has at least one inflection point POI where a concavity or curvature of the resistivity ρ changes along the axis, as shown in the graph of Fig. 2 is illustrated.

[0058] According to one embodiment, the net n-type doping in the silicon ingot is between 1 × 10 13 cm -3 and 1 × 10 15 cm -3 .

[0059] Although specific embodiments are illustrated and described here, it is obvious to those skilled in the art that a multitude of alternatives and / or equivalent designs can be used for the specific embodiments shown and described without departing from the scope of the present invention. This application is intended to cover any adaptations or modifications of the specific embodiments discussed herein. Therefore, it is intended that this invention is limited only by the claims and their equivalents. Reference symbol list d depth I2 2 , I3 2 inclined implantation I1 2 non-angled implantation 100 CZ growth system 105 crucibles 106 crucible carriers 107 Heating system 108 Bearing shaft 109 Supply device 110 Silicon melt 112 silicon ingots 114 Seed crystal 115 germ carriers 116 Draw shaft 125 Dopant source material 126 surface 127 Puller mechanism 128 Control mechanism

Claims

A Czochralski, CZ-growth method of a silicon ingot (112) for IGBTs, the method comprising: melting a mixture of a silicon material and an n-type dopant material in a crucible (105), extracting the silicon ingot (112) from the molten mixture during an extraction period by CZ growth, thereby doping the silicon ingot (112) with the n-type dopant material of the molten mixture, and doping the silicon ingot (112) with an additional n-type dopant material during at least one sub-period of the extraction period, wherein the doping of the silicon ingot (112) with the additional n-type dopant material is in addition to the doping with the n-type dopant material of the molten mixture, and doping the silicon ingot (112) with the additional n-type The dopant material begins after the growth of a first ingot part and ends before the growth of a further ingot part.wherein a net n-type doping in the silicon ingot (112) is set between 1 × 1013cm-3 and 1 × 1015cm-3. The method of claim 1, wherein the additional n-type dopant material is phosphorus. Method according to claim 1 or 2, wherein the silicon ingot (112) is doped with the additional n-type dopant material by a vapor phase doping technique. Method according to one of claims 1 to 3, further comprising controlling an inlet of a dopant precursor gas into a reaction chamber comprising the silicon ingot (112). Method according to one of claims 1 to 2, wherein doping the silicon ingot (112) with the additional n-type dopant material comprises immersing an n-type dopant source material into the crucible (105). The method of claim 5, wherein the silicon ingot is doped with the additional n-type dopant material by setting a depth of the n-type dopant source material in the molten silicon (110), wherein the n-type dopant source material comprises the additional n-type dopant material. The method of claim 6, wherein adjusting the depth of the n-type dopant source material immersed in the molten silicon (110) in the crucible (105) comprises measuring a weight of the n-type dopant source material. Method according to any one of claims 5 to 7, wherein the n-type dopant source material is in the form of one or more rods. Method according to one of claims 5 to 6, wherein the n-type dopant source material is made of quartz or silicon carbide doped with the additional n-type dopant material. Method according to any one of claims 5 to 9, wherein a concentration profile of the additional n-type dopant material at a depth of the n-type dopant source material has a peak below a surface of the n-type dopant source material. A method according to any one of claims 5 to 10, wherein the introduction of the additional n-type dopant material into the n-type dopant source material is carried out by at least one of the following processes: in-situ doping, plasma deposition or deposition through a surface of the n-type dopant source material, ion implantation through the surface of the n-type dopant source material and diffusion through the surface of the n-type dopant source material. Method according to any one of claims 1 to 11, further comprising changing the degree of doping of the silicon ingot (112) with the additional n-type dopant material during the extraction time between no doping and maximum doping. Method according to any one of claims 1 to 12, further comprising doping the silicon ingot (112) with p-type dopant material by adding the p-type dopant material to the silicon melt (110) using a p-type dopant source material and vapor phase doping technology. Method according to claim 13, wherein the p-type dopant material is at least a material consisting of boron, aluminium and gallium. Method according to any one of claims 1 to 12, further comprising doping the silicon ingot (112) with p-type dopant material by immersing a p-type dopant source material into the crucible (105). The method of claim 15, wherein the silicon ingot (112) is doped with the p-type dopant material by setting a depth of the p-type dopant source material in the molten silicon (110), wherein the p-type dopant source material comprises the p-type dopant material. Method according to claim 15 or 16, wherein adjusting the depth of the p-type dopant source material immersed in the molten silicon (110) in the crucible (105) comprises measuring a weight of the p-type dopant source material. Method according to any one of claims 15 to 17, wherein the p-type dopant source material is in the form of one or more rods. Method according to one of claims 15 to 16, wherein the p-type dopant source material is made of quartz or silicon carbide doped with the p-type dopant material. Method according to one of claims 15 to 19, wherein a concentration profile of the p-type dopant material at a depth of the p-type dopant source material has a peak below a surface of the p-type dopant source material. A method according to any one of claims 15 to 20, wherein the introduction of the p-type dopant material into the p-type dopant source material is carried out by at least one of the following processes: in-situ doping, plasma deposition or deposition through a surface of the p-type dopant source material, ion implantation through the surface of the p-type dopant source material and diffusion through the surface of the p-type dopant source material. Silicon ingot (112) produced by one of the methods according to one of the preceding claims, wherein a specific resistance along an axis between opposite ends of the silicon ingot (112) has at least one inflection point where a concavity of the specific resistance changes along the axis without transitioning to a local maximum or minimum.

Images