Crystal pulling method for reducing head resistivity of heavily arsenic-doped silicon single crystal

Abstract

The invention provides a crystal pulling method for reducing the head resistivity of a heavily arsenic-doped silicon single crystal, and belongs to the technical field of doped silicon single crystal production. Comprising the following steps: S1, sequentially carrying out a primary doping process, a primary temperature testing process, a primary seeding process, a re-melting process, a supplementary doping process, a secondary temperature testing process, a secondary seeding process, a shouldering process and an equal-diameter process; s2, if NG appears in the guiding and releasing process or the equal-diameter process, re-melting the crystal bar and carrying out supplementary doping, and then sequentially carrying out a temperature testing process, a crystal guiding process, a shouldering process and an equal-diameter process; s3, when the sum of the first doping multiple and the supplementary doping multiple reaches the total doping multiple, supplementary doping is stopped. By controlling the doping interval time and the total doping multiple, the head resistivity can be reduced and controlled within the upper limit of the specification, and impurities can be doped more uniformly, so that the crystallization rate is increased, the drawing and discharging times are reduced, and the quality of the drawn monocrystalline silicon is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of doped silicon single crystal production, and particularly relates to a crystal pulling method for reducing the resistivity of the head of a heavily doped arsenic silicon single crystal. Background Art

[0002] One of the key quality control requirements for heavily doped single crystals is resistivity. The resistivity specification consists of an upper limit and a lower limit. Controlling the resistivity distribution of the single crystal completely within the upper and lower limits is the ultimate goal of heavily doped resistivity control. The heavily doped resistivity follows the law of high at the head and low at the tail. Therefore, how to control the head resistance of the single crystal within the upper limit of the specification is one of the control requirements. The common resistivity specifications are distributed between 0 - 0.006 Ω·cm. When the upper limit of the specification is between 0.003 - 0.006 Ω·cm, the existing conventional control methods can completely control the head resistance of the single crystal within the upper limit of the specification. However, when the upper limit of the specification < 0.003 Ω·cm, the existing conventional control methods cannot control the head resistivity of the single crystal within the upper limit of the specification, and can only control it within the upper limit of the specification at a certain length after the head, resulting in a loss in yield. This loss is more obvious when the upper limit of the specification is lower. Therefore, there is an urgent need for a control method to reduce the head resistivity to meet the requirements. Summary of the Invention

[0003] In view of this, the present invention provides a crystal pulling method for reducing the resistivity of the head of a heavily doped arsenic silicon single crystal to solve the technical problem that when the upper limit of the specification < 0.003 Ω·cm, the existing crystal pulling methods cannot control the head resistivity within the upper limit of the specification.

[0004] The technical solution adopted by the present invention to solve its technical problems is as follows: A crystal pulling method for reducing the resistivity of the head of a heavily doped arsenic silicon single crystal, comprising the following steps: S1. Sequentially perform the first doping process, the first temperature measurement process, the first crystal seeding process, the remelting process, the supplementary doping process, the second temperature measurement process, the second crystal seeding process, the shoulder releasing process, and the equal diameter process; S2. If NG occurs during the crystal pulling or equal diameter process, remelt the crystal rod and perform supplementary doping, and then sequentially perform the temperature measurement process, the crystal seeding process, the shoulder releasing process, and the equal diameter process; S3. When the sum of the first doping multiple and the supplementary doping multiple reaches the total doping multiple, stop supplementary doping.

[0005] Preferably, the total doping multiple is determined according to the upper limit of the resistivity specification.

[0006] Preferably, the total doping multiple is the sum of the first doping multiple and a variable, where the variable is greater than 2.

[0007] Preferably, the time of both the first temperature measurement process and the second temperature measurement process is ≥ 4H.

[0008] Preferably, step S3 includes the following steps: S31. If NG occurs again during the drawing or equal-diameter process, determine whether the sum of the first doping multiple and the supplementary doping multiple reaches the total doping multiple; S32. If the sum of the first doping multiple and the supplementary doping multiple does not reach the total doping multiple, return to execute step S2; S33. If the sum of the first doping multiple and the supplementary doping multiple reaches the total doping multiple, stop supplementary doping.

[0009] Preferably, after step S33, the method further includes: remelting the ingot, without performing supplementary doping, and successively performing the temperature testing process, crystal seeding process, shoulder forming process, and equal-diameter process.

[0010] Preferably, the time for the temperature testing process after not performing supplementary doping is ≤2H.

[0011] Preferably, if NG does not occur during the drawing or equal-diameter process, the crystal pulling ends.

[0012] Preferably, the supplementary doping amount is the time from the previous supplementary doping process to the next entry into the equal-diameter process * the preset doping amount per hour.

[0013] Compared with the prior art, the beneficial effects of the present invention are as follows: The present invention adds a temperature testing process, a crystal seeding process, a remelting process, and a supplementary doping process, thereby increasing the doping interval time, making the impurities more uniform, and reducing the resistivity of the head. At the same time, the present invention proposes the concept of the total doping multiple. When the total doping multiple is set, supplementary doping is performed. Compared with continuous supplementary doping, it can ensure that the resistivity of the head passes through the lowest point, preventing the resistivity of the head from increasing due to continuous doping. Therefore, by using the crystal pulling method of the present invention, by controlling the doping interval time and the total doping multiple, not only can the resistivity of the head be reduced, and when the specification upper limit <0.003 Ω·cm, the resistivity of the head can be controlled within the specification upper limit, but also the impurity doping can be made more uniform, the crystal formation rate can be increased, the drawing and releasing times can be reduced, and the quality of the drawn single crystal can be guaranteed. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 is a process flow chart of the crystal pulling method for reducing the resistivity of the heavily doped arsenic head of the present invention.

[0015] Figure 2 is a relationship diagram between the resistivity of the head and the total doping multiple.

[0016] Figure 3 is a comparison diagram of the resistivity of the head between Example 1 and Comparative Example A.

[0017] Figure 4Resistivity comparison diagram of the head part between Example 2 and Comparative Example B.

[0018] Figure 5 Resistivity comparison diagram of the head part between Example 3 and Comparative Example C. Detailed implementation manners

[0019] The technical solutions and technical effects of the embodiments of the present invention will be further elaborated in detail below in conjunction with the drawings of the present invention.

[0020] Please refer to Figure 1 , a crystal pulling method for reducing the resistivity of the head part of heavily doped arsenic silicon single crystal, comprising the following steps: S1. Perform the first doping process, the first temperature measurement process, the first crystal seeding process, the remelting process, the supplementary doping process, the second temperature measurement process, the second crystal seeding process, the shoulder releasing process, and the equal diameter process in sequence; S2. If NG occurs during the crystal seeding and releasing process or the equal diameter process, remelt the crystal rod and perform supplementary doping, and then perform the temperature measurement process, the crystal seeding process, the shoulder releasing process, and the equal diameter process in sequence; wherein, NG refers to the destruction of the single crystal structure, such as crystal crack, dislocation, diameter fluctuation, etc. When NG occurs during the crystal seeding and releasing or equal diameter process, it is necessary to remelt the crystal rod for supplementary doping and perform crystal pulling again to ensure the quality of the crystal rod; S3. Stop supplementary doping when the sum of the first doping multiple and the supplementary doping multiple reaches the total doping multiple. Among them, the doping multiple is the ratio of the doping amount to the feeding amount. Therefore, the first doping multiple is the ratio of the first doping amount to the feeding amount, the supplementary doping multiple is the ratio of the supplementary doping amount to the feeding amount, and the total doping multiple is the ratio of the total doping amount (the first doping amount + the supplementary doping amount) to the feeding amount.

[0021] Compared with the traditional crystal pulling process flow, the present invention adds a first temperature measurement process, a first crystal seeding process, a remelting process and a supplementary doping process, thereby increasing the doping interval time, making the impurities more uniform, and reducing the resistivity of the head part. At the same time, the present invention proposes the concept of the total doping multiple. Please refer to Figure 2 , the total doping multiple has an inverse parabolic correlation with the resistivity of the head part, that is, there is a total doping multiple that makes the resistivity of the head part at the lowest point. If doping continues after passing the lowest point, the resistivity of the head part will instead increase. Therefore, when performing supplementary doping under the condition of setting the total doping multiple, compared with continuous supplementary doping, it can be ensured that the resistivity of the head part is past the lowest point, preventing the resistivity of the head part from increasing due to continuous doping. Therefore, by using the crystal pulling method of the present invention, by controlling the doping interval time and the total doping multiple, not only can the resistivity of the head part be reduced, and when the upper limit of the specification is <0.003 Ω·cm, the resistivity of the head part can be controlled within the upper limit of the specification, but also the impurity doping can be made more uniform, thereby improving the crystal formation rate, reducing the number of crystal seeding and releasing times, and ensuring the quality of the pulled single crystal silicon.

[0022] Furthermore, heavily doped substances have volatile properties, and both volatilization and segregation act on the resistivity simultaneously. During the doping process, arsenic vaporizes as the temperature rises, most of it is carried away by argon, and a small part enters the silicon solution. This small part that enters will also volatilize over time. In the low-resistance field, due to conventional control reasons, the pressure and argon settings are relatively high. In an environment of high pressure and high argon, volatilization is inhibited, so over time, the volatilization effect is not obvious. At the same time, after increasing the doping interval time, that is, increasing the effective impurity content in the melt, it also makes the impurities more uniform. Under the combined action of inhibiting volatilization and uniformization, the resistivity of the head is reduced. Moreover, arsenic has the concept of solid solubility in the silicon solution, that is, the silicon solution can only accommodate a certain amount of arsenic impurities. After exceeding this limit, the excess arsenic impurities can basically only be volatilized. A part of the volatilized arsenic impurities will be carried by argon to the surface of the melt, increasing the number of drawing operations. By setting the total doping multiple, on the one hand, it can ensure that the resistivity of the head reaches the lowest point, and on the other hand, it can also reduce the problem of a high number of drawing operations caused by an increase in the amount of volatilized arsenic impurities due to excessive doping.

[0023] At the same time, the newly added process in the present invention includes a supplementary doping process. In the case of single crystal growth, due to the supplementary doping, the total doping amount is more than the first doping amount. Therefore, compared with the traditional crystal pulling process in the case of single crystal growth, the doping amount is more. Thus, compared with the traditional crystal pulling process, the resistivity of the head can be reduced, that is, through the newly added supplementary doping process of the present invention, the resistivity of the head can be reduced even in the case of single crystal growth.

[0024] During the crystal pulling process, the feeding amount is fixed, and the doping amount gradually increases. Therefore, like the doping multiple, the doping amount is inversely parabolic related to the resistivity of the head. However, in processes with the same upper specification limit but different sizes, the base number of the feeding amount is different. If the total doping amount is used for limitation, different feeding amounts will correspond to different total doping amounts. By using the total doping multiple for limitation, in processes with the same upper specification limit but different sizes, only the same total doping multiple needs to be selected, so that the doping amount can be better controlled during the supplementary doping process. In some embodiments, the unit of the doping multiple is g / kg.

[0025] Furthermore, the total doping multiple is determined according to the upper specification limit of the resistivity. Through data statistics, it is found that the total doping multiple and the resistivity of the head show an inversely parabolic correlation, that is, when the total doping multiple is greater than a certain critical value, there is a lowest value for the resistivity of the head. That is, the total doping multiple and the resistivity also show an inversely parabolic correlation. Therefore, to control the resistivity of the head within the upper specification limit of the resistivity, an appropriate total doping multiple can be selected according to the upper specification limit of the resistivity.

[0026] In some embodiments, according to data statistics, the corresponding relationship between the total doping multiple and the resistivity can be determined. Based on the upper limit of the target resistivity specification and the corresponding relationship, the total doping multiple is calculated. Then, additional doping is performed according to the calculated total doping multiple to control the resistivity of the head of the drawn silicon single crystal within the upper limit of the target resistivity specification. Additionally, according to data statistics, a corresponding table of the total doping multiple and the resistivity can be made. Based on the upper limit of the target resistivity specification, the total doping multiple corresponding to this specification upper limit is found from the corresponding table, and then additional doping is performed according to the found total doping multiple.

[0027] In some embodiments, since the corresponding relationship between the total doping multiple and the head resistivity is determined based on the results of data statistics, and data statistics cannot fully cover all head resistivities, in order to improve the accuracy of the corresponding relationship between the total doping multiple and the head resistivity and more precisely control the head resistivity of the heavily doped arsenic silicon single crystal, the corresponding relationship between the total doping multiple and the head resistivity can also be determined through a model. Specifically, there is an inverse parabolic correlation between the total doping multiple and the head resistivity. A polynomial regression model can be established according to the characteristics of the inverse parabola. Then, the statistical data is divided into a training set and a validation set, and the model is trained and fitted. Finally, the corresponding relationship model between the total doping multiple and the head resistivity is obtained. When the upper limit of the resistivity specification for heavily doped arsenic is determined, the specification upper limit can be input into the above corresponding relationship model to obtain the total doping multiple corresponding to the specification upper limit, and then doping is performed using this total doping multiple to obtain the head resistivity of heavily doped arsenic that meets the given specification upper limit.

[0028] Furthermore, the total doping multiple is the sum of the first doping multiple and a variable, where the variable is greater than 2. There is a significant correlation between the doping amount and the head resistivity. The more the doping amount, the lower the head resistivity. However, when the doping amount is sufficient, the number of drawing and feeding operations of the single crystal will increase. After each drawing and feeding, additional doping is required, and each additional doping will increase the oxides in the furnace cavity, resulting in difficulty in crystal formation. Therefore, in the low-resistance field (where the specification upper limit requirement is ≤ 0.003 Ω·cm), a maximum first doping amount needs to be adopted. Therefore, when determining the total doping multiple of this application, a maximum first doping multiple also needs to be determined first. The first doping multiple varies depending on the doping process, but still follows the rule that the lower the required head resistance, the higher the first doping multiple. A reasonable setting of the first doping multiple can achieve the purpose of reducing the number of drawing and feeding operations while meeting the low head resistivity requirement. The size of the variable depends on the amount of the first doping multiple and the size of the specification upper limit. Since the heavily doped arsenic silicon single crystal itself differentiates different processes, different process parameters are distinguished according to the resistivity level and grade, and each process has a first doping multiple and a corresponding variable.

[0029] Furthermore, the time for both the first temperature testing process and the second temperature testing process is ≥ 4H. In the existing process flow, the seeding time is generally < 2.6H, the shoulder releasing time is generally < 3.8H, and the sum of the two times is < 6.4H. The conventional control method requires the doping interval time to be controlled within ≤ 8H. Since the time of other processes cannot be reduced, only the time for temperature testing can be reduced. Therefore, in the conventional control method, the temperature testing time is generally ≤ 2H. However, the purpose of temperature testing is to improve the uniformity of the impurities doped in the melt through a certain period of time. If the time is short, the distribution of impurities is uneven, and the longer the temperature testing time, the more uniform the impurity distribution. At the same time, the uniformity of impurity distribution is also related to the number of NG times. The more uniform the impurity distribution, the fewer the number of NG times. Since the present invention is applied to the low-resistance field, the doping interval time in the low-resistance field can be not limited by the doping interval time of the conventional control method. Therefore, the present invention can increase the time of the temperature testing process, that is, set the time of the temperature testing process to ≥ 4H, so as to improve the uniformity of impurity doping in the melt.

[0030] Furthermore, step S3 includes the following steps: S31. If NG occurs again during the seeding and shoulder releasing process or the equal diameter process, determine whether the sum of the first doping multiple and the supplementary doping multiple reaches the total doping multiple; S32. If the sum of the first doping multiple and the supplementary doping multiple does not reach the total doping multiple, return to execute step S2; S33. If the sum of the first doping multiple and the supplementary doping multiple reaches the total doping multiple, stop supplementary doping.

[0031] In the traditional crystal pulling process, if there is one NG during the feeding process or the constant diameter process, one additional doping is required, and then this process is repeated. However, in this application, the judgment of the total doping multiple is added during this process, so as to avoid the increase in resistivity caused by continuous additional doping by limiting the total doping multiple. Specifically, after step S2, it is necessary to first determine whether there is NG again during the feeding process or the constant diameter process. If there is no NG, it indicates that the crystal bar has met the process requirements and can enter the finishing stage to end the crystal pulling. Then there is no need for additional doping, and there is no need to judge the total doping multiple because the crystal pulling has been completed. If there is NG again during the feeding process or the constant diameter process, additional doping is required, but before the additional doping, the judgment of the total doping multiple is required. If the sum of the first doping multiple and the additional doping multiple does not reach the total doping multiple, that is, the sum of the first doping multiple and the additional doping multiple is less than the total doping multiple, then step S2 can be returned to perform remelting and additional doping on the crystal bar, and then continue to perform the temperature testing process, the crystal seeding process, the shoulder forming process, and the constant diameter process in sequence, and then judge whether there is NG during the feeding process or the constant diameter process, that is, repeat the above process. However, if the sum of the first doping multiple and the additional doping multiple reaches the total doping multiple, that is, the sum of the first doping multiple and the additional doping multiple is greater than or equal to the total doping multiple, since there is an inverse parabolic correlation between the total doping multiple and the head resistivity, the set total doping multiple corresponds to the lowest head resistivity. When the sum of the first doping multiple and the additional doping multiple is greater than or equal to the total doping multiple, if additional doping continues, it will instead increase the head resistivity of the silicon single crystal. Therefore, at this time, additional doping should be stopped to keep the head resistivity of the silicon single crystal at the lowest point, and then continue with the temperature testing, crystal seeding, shoulder forming, and constant diameter processes to ensure the quality of the silicon single crystal.

[0032] Furthermore, after step S33, the method further includes: remelting the crystal bar without additional doping, and performing the temperature testing process, the crystal seeding process, the shoulder forming process, and the constant diameter process in sequence. Since the temperature testing process, the crystal seeding process, the shoulder forming process, and the constant diameter process are still performed after the additional doping, if there is NG again during the feeding process or the constant diameter process, the crystal bar needs to be remelted again to perform crystal pulling. However, since the sum of the first doping multiple and the additional doping multiple has reached the total doping multiple, if additional doping continues, it will increase the head resistivity of the crystal bar. Therefore, additional doping should be stopped and the temperature testing process, the crystal seeding process, the shoulder forming process, and the constant diameter process should be continued. If there is NG again during this process during the feeding process or the constant diameter process, the crystal bar still needs to be remelted, but without additional doping, and continue with the temperature testing process, the crystal seeding process, the shoulder forming process, and the constant diameter process. Repeat this process until there is no NG during the feeding process or the constant diameter process, then the crystal pulling can be ended to obtain a well-drawn heavily doped arsenic silicon single crystal. In this way, both the head resistivity of the heavily doped arsenic silicon single crystal can be reduced and the quality of the crystal bar can be ensured.

[0033] Further, the time for the temperature testing process without additional doping is ≤2H. After stopping the additional doping, if NG occurs during the seed pulling or equal diameter process, the ingot needs to be remelted without additional doping, and then the temperature testing process, seed crystal pulling process, shoulder forming process, and equal diameter process are carried out again until NG does not occur during the seed pulling or equal diameter process, and then the crystal pulling ends. In the previous process with additional doping, the time for the temperature testing process should be ≥4H to ensure that the doped arsenic can be evenly incorporated into the silicon melt. However, since additional doping has now stopped and there is no need to increase the time for the temperature testing process to ensure the uniformity of impurity doping, the time for the temperature testing process can be adjusted to ≤2H, thereby reducing the crystal pulling time and improving the crystal pulling efficiency.

[0034] Further, if NG does not occur during the seed pulling or equal diameter process, the crystal pulling ends. During the entire crystal pulling process, as long as NG does not occur during the seed pulling or equal diameter process, that is, there are no crystal cracks, dislocations, diameter fluctuations, etc., indicating that the crystal structure is complete and the parameter control is effective, and the process requirements have been met, then the finishing stage can be carried out to end the crystal pulling. Even if NG does not occur during the seed pulling or equal diameter process after performing step S1, that is, one-time crystal formation, the crystal pulling also directly ends without the need for additional doping or judgment of the total doping multiple. And since in step S1, compared with the traditional crystal pulling process, a new temperature testing process, a new seed crystal pulling process, a remelting process, and an additional doping process are added, and additional doping will reduce the resistivity of the head of the ingot, therefore, even for one-time crystal formation, based on the new process of the present application, compared with the traditional crystal pulling process, the resistivity of the head of the ingot can still be reduced.

[0035] Further, the additional doping amount is the time from the previous additional doping process to the next equal diameter process * the preset doping amount per hour. Among them, the preset doping amount per hour is determined according to the upper specification limit and doping parameters, and the unit of the doping amount is grams.

[0036] The following specific experimental examples are used to further illustrate the technical solutions and technical effects of the present invention. It should be noted that the following experimental examples are only for further explaining the present invention and do not limit the technical solutions of the present invention.

[0037] Test settings: Divided into two groups (control group and experimental group), the control group is controlled by the original conventional method, and the experimental group is the crystal pulling method of the present invention. Among them, the following targets are set: Upper specification limit: 0.0022Ω·cm. The test is divided into two cases of NG 0 times and NG 3 times. The crystal pulling parameters of the experimental group and the control group follow the same settings, including crystal rotation: 12rpmm; crucible rotation: 2rpm; pressure: 30Kpa; argon: 120slm; doping device: multi-layer quartz bell jar; first doping multiple: 8.0; MCZ; Gap: 30mm.

[0038] Parameter settings for the comparative examples (NG0 times and NG3 times): Head resistance: 0.0022 Ω·cm; Controlled by conventional methods, the process flow is: doping - temperature testing - crystal seeding - shoulder releasing - equal diameter; There is no requirement for the total doping multiple, and the additional doping rule is: doping is carried out at a rate of 8 g / H for the time from the previous additional doping to the next entry into the equal-diameter stage. Among them, the process flow and process time of the comparative examples are shown in Table 1.

[0039] Table 1

[0040] Parameter settings for the examples (NG0 times and NG3 times): Head resistance: 0.0022 Ω·cm; Controlled by the method of the present invention, the process flow is: first doping - first temperature testing - first crystal seeding - remelting - additional doping - second temperature testing - second crystal seeding - shoulder releasing - equal diameter; The total doping multiple is 10.0, and the additional doping rule is: doping is carried out at a rate of 8 g / H for the time from the previous additional doping to the next entry into the equal-diameter stage. Among them, the process flow and process time of the examples are shown in Table 2.

[0041] Table 2

[0042] When detecting the head resistivity of the comparative example and the example in the case of NG0 times, the head resistivity comparison graph as shown in Figure 3 is obtained. Among them, Example 1 is the head resistivity of the ingot controlled by the method of the present invention in the case of NG0 times, and Comparative Example A is the head resistivity of the ingot controlled by the conventional method in the case of NG0 times. From Figure 3 it can be seen that the head resistivity of Example 1 is mainly distributed between 0.0025 - 0.0026 Ω·cm, and the head resistivity of Comparative Example A is mainly distributed between 0.0027 - 0.0028 Ω·cm. The head resistivity of Example 1 is lower than that of Comparative Example A. Since the total doping multiple is not involved in the case of NG0 times, the reason for the lower head resistivity of Example 1 is that the present invention adds one temperature testing - one crystal seeding - remelting - additional doping compared with the traditional process. On the one hand, it increases the doping interval time, which can make the doped impurities more uniform and reduce the head resistivity. On the other hand, it adds the additional doping process. In the case of one-time crystal formation, that is, in the case of NG0 times, the original doping amount can also be increased based on the additional doping, thereby reducing the resistivity even in the case of one-time crystal formation.

[0043] When detecting the head resistivity of the comparative example and the example in the case of NG3 times, the head resistivity comparison graph as shown in Figure 4 is obtained. Among them, Example 2 is the head resistivity of the ingot controlled by the method of the present invention in the case of NG3 times, and Comparative Example B is the head resistivity of the ingot controlled by the conventional method in the case of NG3 times. FromFigure 4 It can be seen that the head resistivity of Example 2 is mainly distributed around 0.0024Ω·cm, and the head resistivity of Comparative Example B is mainly distributed between 0.00275-0.0028Ω·cm. The head resistivity of Example 2 is significantly lower than that of Comparative Example B.

[0044] First, compare Figure 3 , Figure 4 The distribution and mean of the head resistivity of Comparative Example A and Comparative Example B show that after three times of supplementary doping, the head resistivity of Comparative Example B is higher than that of Comparative Example A. In theory, the more doping, the lower the resistivity. Therefore, this shows that in the process of continuous supplementary doping, the head resistivity will not continue to decrease with the continuous increase of doping, but will increase after exceeding a certain critical value. This also confirms that the head resistivity and doping amount are indeed inversely parabolic. Figure 3 , Figure 4 From the distribution and mean of the head resistivity of Example 1 and Example 2, it can be seen that the head resistivity of Example 2 is lower than that of Example 1. It can be seen from Table 2 that the doping multiple has reached the set total doping multiple at NG2, so no additional doping is performed in the next process. Combined with the comparison results of Comparative Example A and Comparative Example B, it can be explained that after setting the total doping multiple, the head resistivity can be ensured to be at the lowest point, thereby preventing the head resistivity from increasing due to continuous doping. Finally, by comparing Figure 3 The head resistivity of Example 1 and Comparative Example A in the embodiment 1 and the comparative example A can illustrate that the newly added process of the present invention can reduce the head resistivity of the heavily arsenic-doped silicon single crystal. Figure 3 , Figure 4 The head resistivity of Example 1 and Example 2 can illustrate that setting the total doping multiple can reduce the head resistivity of heavily arsenic-doped silicon single crystal. Figure 4 The head resistivity of Example 2 is higher than that of Comparative Example B due to the combined effect of the newly added process and the set total doping multiple. In summary, the present invention can effectively reduce the head resistivity of heavily arsenic-doped silicon single crystal by adding one temperature test, one seeding, remelting, and supplementary doping processes, as well as setting the total doping multiple.

[0045] In some embodiments, the crystal pulling method of the present invention can be applied not only to heavily doped arsenic silicon single crystals, but also extended to heavily doped red phosphorus silicon single crystals, and can be utilized for both 8-inch and 12-inch crystals. Specifically, please refer to the following experiments. Among them, the test parameters for the 8-inch red phosphorus applying the crystal pulling method of the present invention are set as follows: Target setting: Upper specification limit: 0.0011 Ω·cm, NG 3 times. The rest of the parameters in Example 3 and Comparative Example C follow the same settings, including crystal rotation: 16 rpmm; crucible rotation: 6 rpm; pressure: 25 Kpa; argon: 110 slm; doping apparatus: quartz bell jar; first doping multiple: 9.0; Gap: 25 mm. Example 3

[0046] The process flow is: first doping - first temperature testing - first crystal seeding - remelting - additional doping - second temperature testing - second crystal seeding - shoulder release - equal diameter. The total doping multiple is 10.5. Additional doping rule: Doping is carried out at a rate of 7 g / H based on the time from the previous additional doping to the next entry into the equal diameter stage.

[0047] Comparative Example C Controlled by the conventional method, the process flow is: doping - temperature testing - crystal seeding - shoulder release - equal diameter. There is no requirement for the total doping multiple. Additional doping rule: Doping is carried out at a rate of 7 g / H based on the time from the previous additional doping to the next entry into the equal diameter stage.

[0048] The head resistivities of the silicon single crystals in Example 3 and Comparative Example C are respectively detected, and the obtained comparison chart is as shown in Figure 5 As can be seen from the figure, the head resistivity of the silicon single crystal in Example 3 is between 0.0011 - 0.00115 Ω·cm, while the head resistivity of the silicon single crystal in Comparative Example C is between 0.0012 - 0.0013 Ω·cm. The head resistivity of Example 3 is lower than that of Comparative Example C. Thus, it can be seen that the crystal pulling method of the present invention can also reduce the head resistivity of heavily doped phosphorus silicon single crystals.

[0049] The above-disclosed are only the preferred embodiments of the present invention. Of course, the scope of the rights of the present invention cannot be limited thereby. Those of ordinary skill in the art can understand and implement all or part of the above processes, and the equivalent changes made according to the claims of the present invention still fall within the scope covered by the invention.

Claims

1. A crystal pulling method for reducing the head resistivity of heavily arsenic-doped silicon single crystal, characterized in that: The following steps are involved: S1, sequentially performing the first doping process, the first temperature test process, the first seeding process, the remelting process, the supplementary doping process, the second temperature test process, the second seeding process, the shoulder release process, and the equal diameter process; S2. If NG occurs in the seeding process or the equal diameter process, the crystal rod is melted back and supplemented with doping, and then the temperature test process, seeding process, shoulder release process, and equal diameter process are carried out in sequence; S3. When the sum of the initial doping multiple and the supplementary doping multiple reaches the total doping multiple, stop supplementary doping.

2. The crystal pulling method for reducing the head resistivity of heavily arsenic-doped silicon single crystal according to claim 1, characterized in that: The total doping factor is determined based on the upper specification limit of the resistivity.

3. The crystal pulling method for reducing the head resistivity of heavily arsenic-doped silicon single crystal according to claim 2, characterized in that: The total doping multiple is the sum of the first doping multiple and a variable, wherein the variable is greater than 2.

4. The crystal pulling method for reducing the head resistivity of heavily arsenic-doped silicon single crystal according to claim 1, characterized in that: The time for the first temperature test process and the second temperature test process is ≥4H.

5. The crystal pulling method for reducing the head resistivity of heavily arsenic-doped silicon single crystal according to claim 1, characterized in that: The step S3 comprises the following steps: S31, if NG occurs again in the lead-in process or the equal diameter process, determine whether the sum of the first doping multiple and the supplementary doping multiple reaches the total doping multiple; S32, if the sum of the initial doping multiple and the supplementary doping multiple does not reach the total doping multiple, return to step S2; S33. If the sum of the initial doping multiple and the supplementary doping multiple reaches the total doping multiple, stop supplementary doping.

6. The crystal pulling method for reducing the resistivity of the head of heavily arsenic-doped silicon single crystal according to claim 5, characterized in that: After step S33, the method further includes: remelting the crystal rod without supplementary doping, and sequentially performing a temperature test process, a crystal seeding process, a shoulder release process and a diameter equalization process.

7. The crystal pulling method for reducing the head resistivity of heavily arsenic-doped silicon single crystal according to claim 6, characterized in that: The time of the temperature test process without supplementary doping is ≤2H.

8. The crystal pulling method for reducing the head resistivity of heavily arsenic-doped silicon single crystal according to any one of claims 1 to 7, characterized in that: If no NG occurs during the lead-in process or the equal-diameter process, the crystal pulling is completed.

9. The crystal pulling method for reducing the head resistivity of heavily arsenic-doped silicon single crystal according to claim 1, characterized in that: The amount of supplementary doping is the time from the last supplementary doping process to the next equal diameter process * the preset doping amount per hour.

Images