A method for determining a molecular beam epitaxy growth process

Abstract

The application provides a determination method of a molecular beam epitaxy growth process, and relates to the technical field of semiconductor manufacturing. The method comprises the following steps: selecting a first Ga beam flow rate and a rotation speed R of a sample holder, so that the number of rotations of the sample holder during growth of a channel layer is an integer; selecting a second Ga beam flow rate, so that the number of rotations of the sample holder during growth of a first isolation layer is an integer; selecting a third Ga beam flow rate, so that the number of rotations of the sample holder during growth of a second isolation layer is an integer; and immediately adjusting the source temperature of a second Ga source furnace after the growth of the first isolation layer is completed. According to the application, the number of rotations of the sample holder during the growth of each of the first isolation layer, the channel layer and the second isolation layer is an integer, so that the thickness uniformity during the growth of each layer is ensured. Compared with the optimization of only the rotation speed R, the epitaxial wafer uniformity is further improved, and the performance and yield of the device are improved.

Description

Technical Field

[0001] This invention relates to the field of semiconductor manufacturing technology, and more specifically to a method for determining a molecular beam epitaxy (MBE) growth process. Background Technology

[0002] In large-scale molecular beam epitaxy (MBE) production, multiple substrates are typically supported on a single substrate holder for batch MBE growth, thereby improving production efficiency and reducing production costs. The quality of the intra-wafer uniformity and inter-wafer uniformity of the epitaxial layers on the substrate directly affects the performance and yield of subsequent device fabrication.

[0003] GaAs-based pseudomorphic high electron mobility transistors (pHEMTs) are typically fabricated using molecular beam epitaxy (MBE). During the MBE growth of pHEMT devices, the thickness uniformity of the channel layer and isolation layer affects the doping concentration uniformity of the channel layer. Because the molecular beam velocity emitted from the source furnace (e.g., In furnace, Ga furnace, and Al furnace) providing the molecular beam is not uniformly distributed on the substrate during MBE growth, a conventional method to improve the uniformity of the epitaxial wafer is to rotate the sample holder supporting the substrate at a constant speed during the growth process. Further optimization of the rotation speed ensures that the equivalent beam velocities of various molecular beam sources actually reaching the substrate are relatively uniform over a given period.

[0004] However, the pHEMT material structure comprises multiple critical material layers, each with a growth time that is not necessarily the same. Since the sample holder rotation speed remains constant throughout the entire growth process, optimizing the rotation speed can only improve the uniformity of a subset of critical material layers, potentially degrading the uniformity of others. The uniformity of the epitaxial wafer is the overall effect of the combined properties of all critical material layers; therefore, optimizing the rotation speed alone has limited ability to improve the uniformity of the epitaxial wafer. To further improve the uniformity of the epitaxial wafer, a novel method for determining the epitaxial growth process is needed. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of the prior art by providing a method for determining the molecular beam epitaxy (MBE) growth process, thereby solving the problem of further improving the uniformity of pHEMT epitaxial material growth.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows: This invention provides a method for determining a molecular beam epitaxy (MBE) growth process. This method is used to determine the growth process of a pHEMT epitaxial structure, which includes a first AlGaAs isolation layer, an InGaAs channel layer, and a second AlGaAs isolation layer arranged sequentially from bottom to top. The method includes: For the InGaAs channel layer, a first Ga beam rate and a sample holder rotation speed R are selected such that the number of rotations of the sample holder is an integer during the deposition and growth of the entire InGaAs channel layer, and the rotation speed R remains constant throughout the entire pHEMT epitaxial structure growth. The first Ga beam rate is provided by the first Ga source furnace. For the first AlGaAs isolation layer, a second Ga beam rate is selected such that the number of rotations of the sample holder is an integer during the entire deposition and growth of the first AlGaAs isolation layer. The second Ga beam rate is provided by the second Ga source furnace or by both the first Ga source furnace and the second Ga source furnace. For the second AlGaAs isolation layer, a third Ga beam rate is selected such that the number of rotations of the sample holder is an integer during the entire deposition and growth of the second AlGaAs isolation layer. The Ga source furnaces providing the second and third Ga beam rates are the same. After the growth of the first AlGaAs isolation layer is completed, the difference ΔV between the third and second Ga beam rates is calculated. Ga Immediately adjust the source furnace temperature of the second Ga source furnace, ΔV Ga The following condition must be met: ΔV Ga The time required for the source furnace temperature change of the corresponding second Ga source furnace does not exceed the time required for growing the InGaAs channel layer.

[0007] Optionally, the pHEMT epitaxial structure further includes a first silicon δ-doped layer disposed under the first AlGaAs isolation layer and a second silicon δ-doped layer disposed on the second AlGaAs isolation layer, and the method further includes: The first doping time of the first silicon δ-doped layer and the second doping time of the second silicon δ-doped layer are determined based on the rotation speed R. During the first doping time and the second doping time, the number of rotations of the sample holder are both integers. The silicon source furnace doping temperature of the first silicon δ-doped layer is determined based on the expected doping concentration of the first silicon δ-doped layer and the first doping time. The silicon source furnace doping temperature of the second silicon δ-doped layer is determined based on the expected doping concentration of the second silicon δ-doped layer and the second doping time.

[0008] Optionally, the expected doping concentration of the second silicon δ-doped layer is an integer multiple of the expected doping concentration of the first silicon δ-doped layer, and the silicon source furnace doping temperature of the second silicon δ-doped layer is the same as that of the first silicon δ-doped layer.

[0009] Optionally, the rotational speed R can be in the range of 15 rpm ≤ R ≤ 30 rpm.

[0010] Optionally, the range of the first Ga beam current rate is greater than or equal to 0.1 μm / h and less than or equal to 0.7 μm / h; the range of the second Ga beam current rate is greater than or equal to 0.1 μm / h and less than or equal to 1 μm / h.

[0011] Optionally, the range of the first Ga beam current rate is greater than or equal to 0.2 μm / h and less than or equal to 0.5 μm / h; the range of the second Ga beam current rate is greater than or equal to 0.2 μm / h and less than or equal to 0.7 μm / h.

[0012] Alternatively, the Ga beam rate that is closest to the second Ga beam rate among all Ga beam rates during the deposition and growth of the entire second AlGaAs isolation layer, where the number of rotations of the sample holder is an integer, is taken as the third Ga beam rate.

[0013] The beneficial effects of this invention include: The method for determining the molecular beam epitaxy (MBE) growth process provided by this invention is used to determine the growth process of a pHEMT epitaxial structure. The pHEMT epitaxial structure includes a first AlGaAs isolation layer, an InGaAs channel layer, and a second AlGaAs isolation layer, arranged sequentially from bottom to top. The method includes: for the InGaAs channel layer, selecting a first Ga beam current rate and a sample holder rotation speed R, such that the number of rotations of the sample holder is an integer during the deposition and growth of the entire InGaAs channel layer, and the rotation speed R remains constant throughout the entire pHEMT epitaxial structure growth. The first Ga beam current rate is provided by a first Ga source furnace; for the first AlGaAs… For the isolation layer, a second Ga beam rate is selected such that the number of rotations of the sample holder is an integer during the entire deposition and growth of the first AlGaAs isolation layer. The second Ga beam rate is provided by the second Ga source furnace or by both the first and second Ga source furnaces. For the second AlGaAs isolation layer, a third Ga beam rate is selected such that the number of rotations of the sample holder is an integer during the entire deposition and growth of the second AlGaAs isolation layer. The Ga source furnaces providing the second and third Ga beam rates are the same. After the growth of the first AlGaAs isolation layer is completed, the difference ΔV between the third and second Ga beam rates is calculated. Ga Immediately adjust the source furnace temperature of the second Ga source furnace, ΔV Ga The following condition must be met: ΔV GaThe time required for the temperature change of the corresponding second Ga source furnace does not exceed the time required for growing the InGaAs channel layer. This application first selects the first Ga beam rate and the sample holder rotation speed R based on the InGaAs channel layer. Then, it selects the second Ga beam rate based on the first AlGaAs isolation layer. Finally, based on the second AlGaAs isolation layer, it adjusts the temperature of the second Ga source furnace to achieve the third Ga beam rate. This ensures that the number of rotations of the sample holder is an integer when growing each of the first AlGaAs isolation layer, the InGaAs channel layer, and the second AlGaAs isolation layer, thereby ensuring the thickness uniformity of each layer during growth. Compared to optimizing only the rotation speed R, this further improves the uniformity of the epitaxial wafer, enhancing device performance and yield. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 A schematic diagram of a pHEMT epitaxial structure provided in an embodiment of the present invention is shown; Figure 2 A flowchart illustrating a method for determining a molecular beam epitaxy growth process according to an embodiment of the present invention is shown. Figure 3 A schematic diagram of a pHEMT epitaxial structure provided in another embodiment of the present invention is shown. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] For molecular beam epitaxy (MBE) growth of pHEMT epitaxial structures, optimizing the rotation speed alone has limited ability to improve the uniformity of the epitaxial wafer. To further improve uniformity, a novel method for determining the epitaxial growth process is needed.

[0018] For any source furnace in a molecular beam epitaxy system, the beam rate is non-uniform at different positions on the substrate tray. When the substrate tray is stationary, the beam rate is greater closer to the source furnace. Therefore, the uniformity of the deposition rate at different positions on the substrate tray can be improved by rotating the substrate tray. Taking a Ga source furnace as an example, the Ga beam rate at different positions on the substrate tray satisfies the following relationship: .

[0019] in Vr Let α represent the Ga beam current rate, and θ be a coefficient obtained by fitting the growth rate data of the entire substrate tray. For a fixed molecular beam epitaxy (MBE) device, α is a fixed value and can be obtained in advance through testing. For different source furnaces of the same device, the value of α can be considered equal. r represents the distance from the center of the substrate tray, θ represents the azimuth angle, and θ1 represents the phase shift based on the position of the Ga source furnace. When the position of the Ga source furnace is determined, θ1 is a fixed value that can be obtained in advance. R0 represents the rotation speed of the substrate tray (i.e., the rotation speed of the sample holder), in revolutions per minute (rpm); t represents time, in minutes; and Vr0 represents the deposition rate at the center point of the substrate tray. Integrating the above relationship over time t yields the deposition thickness distribution at different positions on the substrate tray within a predetermined time period. For example, for the time period from time t1 to time t2, integrating the above relationship yields the deposition thickness of the Ga source furnace at different positions on the substrate tray from time t1 to time t2: .

[0020] Applying the trigonometric function and difference-to-product formulas to the above equation, it can be transformed into: .

[0021] As can be seen from the above formula, when When the value of is equal to an integer, the above formula is further simplified to: This indicates that the deposition thickness of the source furnace is independent of r and θ. In other words, the deposition thickness of the source furnace is consistent across the entire substrate, resulting in optimal uniformity. The value being an integer indicates that the number of rotations of the sample holder within a preset time period is an integer. In other words, to achieve optimal uniformity of a specified layer during deposition, the number of rotations of the sample holder within the growth time period should be an integer. However, in actual molecular beam epitaxy (MBE), once the sample holder rotation speed is determined, it remains constant throughout the growth of each layer in the entire device structure. The thickness of each layer in the device structure is a pre-designed fixed value. For each source furnace required, after selecting a suitable beam rate and keeping it constant, the growth time required for each layer can be calculated. With a fixed sample holder rotation speed, it is difficult to ensure that the number of rotations of the sample holder is exactly an integer within the growth time period for each layer. Therefore, optimizing the rotation speed alone has limited ability to improve the uniformity of the epitaxial wafer.

[0022] Therefore, it is necessary to design a method for determining the growth process. For the pHEMT device material structure, it is desirable that the number of rotations of the sample holder is an integer when growing each critical material layer.

[0023] This invention provides a method for determining the molecular beam epitaxy (MBE) growth process, which is used to determine the growth process of pHEMT epitaxial structures. For example... Figure 1 As shown, the pHEMT epitaxial structure includes, from bottom to top, a GaAs substrate 10, a buffer layer 11, a first AlGaAs isolation layer 12, an InGaAs channel layer 13, and a second AlGaAs isolation layer 14. The key material layers in this epitaxial structure include the first AlGaAs isolation layer 12, the InGaAs channel layer 13, and the second AlGaAs isolation layer 14. It should be understood that a complete pHEMT epitaxial structure may also include, in addition to... Figure 1 Other epitaxial layers besides the epitaxial layer shown.

[0024] The purpose of the method provided in this embodiment is to ensure that the number of rotations of the sample holder is an integer when growing each of the first AlGaAs isolation layer 12, the InGaAs channel layer 13, and the second AlGaAs isolation layer 14.

[0025] Figure 2 A flowchart illustrating a method for determining a molecular beam epitaxy growth process according to an embodiment of the present invention is shown, as follows. Figure 2 As shown, the method includes: Step 201: For the InGaAs channel layer, select the first Ga beam current rate and the rotation speed R of the sample holder, so that the number of rotations of the sample holder during the deposition and growth of the entire InGaAs channel layer is an integer.

[0026] The first Ga beam rate is used to provide the Ga beam required for InGaAs channel layer growth.

[0027] The thicknesses of the first AlGaAs isolation layer 12, the InGaAs channel layer 13, and the second AlGaAs isolation layer 14 are all pre-designed fixed thicknesses. Since the thickness of the InGaAs channel layer 13 is fixed, an initial value V is selected for the growth rate (the sum of the In beam rate and the Ga beam rate) of the InGaAs channel layer 13. 13 Therefore, the time t required to grow the InGaAs channel layer 13 can be calculated. 13 Then, a rotational speed R is selected in the molecular beam epitaxy device within an empirically selectable rotational speed range, such that time t 13 Within this range, the number of rotations is closest to an integer. Optionally, the rotational speed R can range from 15 revolutions per minute to R ≤ 30 revolutions per minute. It should be noted that in this application, all values ​​of the rotational speed R are integers. If time t... 13 The number of internal rotations is already an integer, based on the initial value V. 13 By determining the In composition in the InGaAs channel layer 13, the corresponding Ga beam rate can be calculated, and this Ga beam rate can be used as the first Ga beam rate. If time t... 13 Within a given time t, if the selected rotational speed corresponds to a number of revolutions that is close to an integer but not an integer, then the rotational speed R is fixed. 13 The time t' is obtained by making the minimum adjustment. 13 So that at time t' 13 Within, the number of rotations is an integer, then based on time t' 13 Based on the thickness of the InGaAs channel layer 13, the growth rate V' of the InGaAs channel layer 13 was recalculated. 13 Then, based on the In composition in the InGaAs channel layer 13, the corresponding Ga beam rate can be calculated, and this Ga beam rate is taken as the first Ga beam rate. For time t... 13 The time t' is obtained by making the minimum adjustment. 13 Specifically, this involves rounding the number of revolutions corresponding to the selected rotation speed to an integer, and then dividing the integer by the rotation speed R to obtain the adjusted time t'. 13 Specifically, for example, the InGaAs channel layer 13 has a thickness of 8 nm, an In composition of 0.25, and an initial V value. 13 If the speed is set to 0.4 μm / h, then the calculated time t is... 13The time is 72 seconds. If the rotation speed R is chosen to be 25 revolutions per minute, then the total number of rotations during the growth of the InGaAs channel layer 13 is exactly 30. At this point, the first Ga beam current rate is calculated to be 0.3 μm / h. Therefore, a first Ga beam current rate of 0.3 μm / h and a rotation speed R of 25 revolutions per minute can be chosen as the final selection. For example, the thickness of the InGaAs channel layer 13 is 7.5 nm, the In composition is 0.25, and the initial value V... 13 If the speed is set to 0.4 μm / h, then the calculated time t is... 13 The time is 67.5 seconds. At this point, with R equal to 25 revolutions per minute, the calculated number of rotations is 28.125, which is not an integer. Rounding to 28, the time corresponding to 28 rotations is 67.2 seconds. (This is achieved by using t...) 13 Adjust to obtain time t' 13 =67.2 seconds, then according to t' 13 =67.2 seconds and a thickness of 7.5 nm, the growth rate V' of the InGaAs channel layer 13 was recalculated. 13 With an In composition of 0.25, the calculated first Ga beam rate is approximately 0.3013 μm / h. Therefore, the first Ga beam rate of 0.3013 μm / h and the rotation speed R of 25 rpm can be used as the final selection of process parameters for depositing and growing the entire InGaAs channel layer in step 201.

[0028] When selecting the growth rate (the sum of the In beam rate and the Ga beam rate) of the InGaAs channel layer 13, it is necessary to consider the preset In composition and the empirical range of the first Ga beam rate. While ensuring a certain growth rate, it is important to avoid excessively high Ga source furnace temperatures. It should be noted that the first Ga beam rate is provided by the first Ga source furnace. Optionally, the range of the first Ga beam rate is greater than or equal to 0.1 μm / h and less than or equal to 0.7 μm / h. Optionally, the range of the first Ga beam rate is greater than or equal to 0.2 μm / h and less than or equal to 0.5 μm / h.

[0029] It should be understood that once the rotational speed R is selected for the InGaAs channel layer, the rotational speed R remains constant throughout the entire pHEMT epitaxial structure growth.

[0030] Step 202: For the first AlGaAs isolation layer, select the second Ga beam rate such that the number of rotations of the sample holder during the entire deposition and growth of the first AlGaAs isolation layer is an integer.

[0031] The second Ga beam current rate is provided by a second Ga source furnace or jointly by both the first and second Ga source furnaces. Whether the second Ga beam current rate is provided by the second Ga source furnace or jointly by both depends on the value of the second Ga beam current rate. If the second Ga beam current rate is much larger than the first Ga beam current rate, to avoid excessive temperature rise in the Ga source furnace when the second Ga beam current rate is provided by only one Ga source furnace, it can be selected that the second Ga beam current rate is provided jointly by both the first and second Ga source furnaces. In this case, the first Ga source furnace still provides the first Ga beam current rate, and the second Ga source furnace provides the difference between the second and first Ga beam current rates. If the second Ga beam current rate is comparable to or less than the first Ga beam current rate, then it can be selected that only the second Ga source furnace provides the second Ga beam current rate. In conventional mass production molecular beam epitaxy (MBE) equipment, it is standard to have two or more Ga source furnaces. Therefore, in conventional mass production MBE equipment, the requirements for the first Ga beam current rate and the second Ga beam current rate described in this application can be met.

[0032] Referring to the calculations in the InGaAs channel layer 13, since the thickness of the first AlGaAs isolation layer 12 is determined, an initial value V is selected for the growth rate (the sum of the Al beam rate and the Ga beam rate) of the first AlGaAs isolation layer 12. 12 Therefore, the time t required to grow the first AlGaAs isolation layer 12 can be calculated. 12 Then, the number of rotations during the growth of the first AlGaAs isolation layer 12 is calculated under the rotation speed R determined in step 201. If the number of rotations is already an integer, it is calculated based on the initial value V. 12 By determining the Al composition in the first AlGaAs isolation layer 12, the corresponding Ga beam rate can be calculated, and this Ga beam rate can be used as the second Ga beam rate. If time t... 12 If the calculated number of rotations is not an integer, then for time t... 12 The time t' is obtained by making the minimum adjustment. 12 So that at time t' 12 Within, the number of rotations is an integer, then based on time t' 12 Based on the thickness of the first AlGaAs isolation layer 12, the growth rate V' of the first AlGaAs isolation layer 12 was recalculated. 12 Then, based on the composition of Al in the first AlGaAs isolation layer 12, the corresponding Ga beam rate can be calculated, and this Ga beam rate can be used as the second Ga beam rate.

[0033] When selecting the growth rate (the sum of the Al beam rate and the Ga beam rate) of the first AlGaAs isolation layer 12, it is necessary to consider the preset Al composition and the empirical range of the second Ga beam rate. While ensuring a certain growth rate, excessively high Ga source furnace temperatures should be avoided. Optionally, the range of the second Ga beam rate is greater than or equal to 0.1 μm / h and less than or equal to 1 μm / h. Optionally, the range of the second Ga beam rate is greater than or equal to 0.2 μm / h and less than or equal to 0.7 μm / h.

[0034] Step 203: For the second AlGaAs isolation layer, select a third Ga beam rate such that the number of rotations of the sample holder is an integer during the entire deposition and growth of the second AlGaAs isolation layer. The Ga source furnace providing the second and third Ga beam rates is the same. After the growth of the first AlGaAs isolation layer is completed, the difference ΔV between the third and second Ga beam rates is calculated. Ga Immediately adjust the source furnace temperature of the second Ga source furnace.

[0035] ΔV Ga The following condition must be met: ΔV Ga The time required for the source furnace temperature change of the corresponding second Ga source furnace does not exceed the time required for growing the InGaAs channel layer.

[0036] Immediately adjusting the source furnace temperature of the second Ga source furnace means that when the second Ga source furnace provides the second Ga beam current rate alone or together with the first Ga source furnace, the corresponding second Ga source furnace temperature is T. Ga2 Immediately after the first AlGaAs isolation layer has grown, T Ga2 Adjusted to T Ga3 For example, the temperature can be varied to T using the maximum temperature variation rate (step mode) allowed by the molecular beam epitaxy system. Ga3 Since this temperature change is typically very small and can be completed in a very short time (less than the growth time of the InGaAs channel layer), the remaining time during the InGaAs channel layer growth process can be used as the rate stabilization time, T Ga3 This is used to indicate the temperature of the second Ga source furnace when the second Ga source furnace provides the third Ga beam rate alone or together with the first Ga source furnace.

[0037] The same Ga source furnace providing the second Ga beam current rate and the third Ga beam current rate specifically means: if the second Ga beam current rate is provided only by the second Ga source furnace, then the third Ga beam current rate is also provided only by the second Ga source furnace; if the second Ga beam current rate is provided jointly by both the first Ga source furnace and the second Ga source furnace, then correspondingly, the third Ga beam current rate is also provided jointly by both the first Ga source furnace and the second Ga source furnace.

[0038] It should be noted that if the Ga beam rate is provided jointly by the first Ga source furnace and the second Ga source furnace during the growth of the first AlGaAs isolation layer 12 and the second AlGaAs isolation layer 14, then the temperature of the first Ga source furnace remains constant during the growth of the three layers: the first AlGaAs isolation layer 12, the InGaAs channel layer 13, and the second AlGaAs isolation layer 14, to provide a stable and constant first Ga beam rate. In conventional structures, the Al composition in the first AlGaAs isolation layer 12 and the second AlGaAs isolation layer 14 is the same. Therefore, in this application, it is considered that the Al composition in the first AlGaAs isolation layer 12 and the second AlGaAs isolation layer 14 is the same.

[0039] When selecting the third Ga beam rate, in order to make ΔV Ga To minimize this, an initial value V can be selected for the growth rate (the sum of the Al beam rate and the Ga beam rate) of the second AlGaAs isolation layer 14. 14 This is equal to the final growth rate of the first AlGaAs isolation layer 12 (as described in step 202 above, under different conditions, the final growth rate of the first AlGaAs isolation layer 12 is equal to V). 12 or V' 12 Therefore, the time t required to grow the second AlGaAs isolation layer 14 can be calculated. 14 Then, the number of rotations during the growth of the second AlGaAs isolation layer 14 is calculated under the rotation speed R determined in step 201. If the number of rotations is already an integer, it is calculated based on the initial value V. 14 By determining the Al composition in the second AlGaAs isolation layer 14, the corresponding Ga beam rate can be calculated, and this Ga beam rate can be used as the third Ga beam rate. If time t... 14 If the calculated number of rotations is not an integer, then for time t... 14 The time t' is obtained by making the minimum adjustment. 14 So that at time t' 14 Within, the number of rotations is an integer, then based on time t' 14 Based on the thickness of the second AlGaAs isolation layer 14, the growth rate V' of the second AlGaAs isolation layer 14 was recalculated. 14 Then, based on the composition of Al in the second AlGaAs isolation layer 14, the corresponding Ga beam rate can be calculated, and this Ga beam rate can be used as the third Ga beam rate.

[0040] If the final growth rate of the second AlGaAs isolation layer 14 is equal to the initial value V 14As mentioned above, since the Al composition in the first AlGaAs isolation layer 12 and the second AlGaAs isolation layer 14 is the same, the third Ga beam current rate is equal to the second Ga beam current rate, ΔV Ga If the final growth rate of the second AlGaAs isolation layer 12 is zero, the second Ga source furnace does not need to be temperature-controlled. Optionally, the Ga beam rate closest to the second Ga beam rate among all Ga beam rates that ensure the sample holder rotates an integer number of times during the deposition and growth of the entire second AlGaAs isolation layer is used as the third Ga beam rate. If the final growth rate of the second AlGaAs isolation layer 14 is equal to the initial value V', then... 14 When the Ga beam rate closest to the second Ga beam rate among all Ga beam rates that cause the sample holder to rotate an integer number of times during the deposition and growth of the entire second AlGaAs isolation layer is taken as the third Ga beam rate, ΔV Ga The value of ΔV is very small. Ga The time required for the temperature change of the corresponding second Ga source furnace is much shorter than the time required for growing the InGaAs channel layer. Thus, by utilizing the time for growing the InGaAs channel layer, the temperature of the second Ga source furnace can be varied and the beam rate stabilized. Finally, before the growth of the second AlGaAs isolation layer 14 begins, the second Ga source furnace can be stably used alone or together with the first Ga source furnace to provide the required third Ga beam rate.

[0041] Through the above process, the sample holder rotation speed R and the first to third Ga beam rates were determined. With the beam rates fixed, the corresponding source furnace temperature can be determined based on the empirical relationship between the source furnace temperature and the beam rate. This ensures that the sample holder rotates an integer number of times in each of the three key material layers of pHEMT, theoretically guaranteeing the uniformity of growth for each key material layer. If a second Ga source furnace temperature variation is required to obtain the third Ga beam rate, the temperature variation and beam rate stabilization can be achieved using the growth time of the InGaAs material layer, thus requiring only two Ga source furnaces to achieve the above objectives.

[0042] like Figure 3As shown, the pHEMT epitaxial structure further includes a first silicon δ-doped layer 15 disposed under the first AlGaAs isolation layer 12 and a second silicon δ-doped layer 16 disposed on the second AlGaAs isolation layer 14. The method further includes: determining a first doping time for the first silicon δ-doped layer and a second doping time for the second silicon δ-doped layer based on a rotation speed R, wherein the number of rotations of the sample holder during the first doping time and the second doping time are both integers; determining the silicon source furnace doping temperature of the first silicon δ-doped layer based on the expected doping concentration of the first silicon δ-doped layer and the first doping time; and determining the silicon source furnace doping temperature of the second silicon δ-doped layer based on the expected doping concentration of the second silicon δ-doped layer and the second doping time. Optionally, the expected doping concentration of the second silicon δ-doped layer is an integer multiple of the expected doping concentration of the first silicon δ-doped layer, and the silicon source furnace doping temperature of the second silicon δ-doped layer is the same as that of the first silicon δ-doped layer. In this case, the second doping time is also an integer multiple of the first doping time.

[0043] In summary, this application first selects the first Ga beam rate and the rotation speed R of the sample holder based on the InGaAs channel layer, then selects the second Ga beam rate based on the first AlGaAs isolation layer, and finally achieves the third Ga beam rate by adjusting the temperature of the second Ga source furnace based on the second AlGaAs isolation layer. This ensures that the number of rotations of the sample holder is an integer when growing each of the first AlGaAs isolation layer, the InGaAs channel layer, and the second AlGaAs isolation layer, thereby ensuring the thickness uniformity of each layer during growth. Compared to optimizing only the rotation speed R, the uniformity of each key material layer reaches the theoretical optimum, further improving the uniformity of the epitaxial wafer and enhancing the device performance and yield.

[0044] The above embodiments are only for illustrating the technical concept and features of the present invention, and are intended to enable those skilled in the art to understand the content of the present invention and implement it. They should not be used to limit the scope of protection of the present invention. All equivalent changes or modifications made in accordance with the spirit and essence of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A method for determining a molecular beam epitaxy (MBE) growth process, characterized in that, The method is used to determine the growth process of a pHEMT epitaxial structure, wherein the pHEMT epitaxial structure includes a first AlGaAs isolation layer, an InGaAs channel layer, and a second AlGaAs isolation layer arranged sequentially from bottom to top, and the method includes: For the InGaAs channel layer, a first Ga beam rate and a sample holder rotation speed R are selected such that the number of rotations of the sample holder is an integer during the deposition and growth of the entire InGaAs channel layer, and the rotation speed R remains constant throughout the entire pHEMT epitaxial structure growth. The first Ga beam rate is provided by the first Ga source furnace. For the first AlGaAs isolation layer, a second Ga beam rate is selected such that the number of rotations of the sample holder is an integer during the entire deposition and growth of the first AlGaAs isolation layer. The second Ga beam rate is provided by the second Ga source furnace or by both the first Ga source furnace and the second Ga source furnace. For the second AlGaAs isolation layer, a third Ga beam rate is selected such that the number of rotations of the sample holder is an integer during the entire deposition and growth of the second AlGaAs isolation layer. The Ga source furnaces providing the second and third Ga beam rates are the same. After the growth of the first AlGaAs isolation layer is completed, the difference ΔV between the third and second Ga beam rates is calculated. Ga Immediately adjust the source furnace temperature of the second Ga source furnace, ΔV Ga The following condition must be met: ΔV Ga The time required for the source furnace temperature change of the corresponding second Ga source furnace does not exceed the time required for growing the InGaAs channel layer.

2. The method for determining the molecular beam epitaxy growth process according to claim 1, characterized in that, The pHEMT epitaxial structure further includes a first silicon δ-doped layer disposed under the first AlGaAs isolation layer and a second silicon δ-doped layer disposed on the second AlGaAs isolation layer; the method further includes: Based on the rotation speed R, the first doping time of the first silicon δ-doped layer and the second doping time of the second silicon δ-doped layer are determined. During the first doping time and the second doping time, the number of rotations of the sample holder are both integers. The silicon source furnace doping temperature of the first silicon δ-doped layer is determined based on the expected doping concentration of the first silicon δ-doped layer and the first doping time. The silicon source furnace doping temperature of the second silicon δ-doped layer is determined based on the expected doping concentration of the second silicon δ-doped layer and the second doping time.

3. The method for determining the molecular beam epitaxy growth process according to claim 2, characterized in that, The expected doping concentration of the second silicon δ-doped layer is an integer multiple of the expected doping concentration of the first silicon δ-doped layer, and the silicon source furnace doping temperature of the second silicon δ-doped layer is the same as that of the first silicon δ-doped layer.

4. The method for determining the molecular beam epitaxy growth process according to claim 1, characterized in that, The rotational speed R ranges from 15 rpm to 30 rpm.

5. The method for determining the molecular beam epitaxy growth process according to claim 1, characterized in that, The first Ga beam current rate ranges from 0.1 μm / h to 0.7 μm / h; the second Ga beam current rate ranges from 0.1 μm / h to 1 μm / h.

6. The method for determining the molecular beam epitaxy growth process according to claim 5, characterized in that, The first Ga beam current rate ranges from 0.2 μm / h to 0.5 μm / h; the second Ga beam current rate ranges from 0.2 μm / h to 0.7 μm / h.

7. The method for determining the molecular beam epitaxy growth process according to claim 1, characterized in that, The Ga beam rate that is closest to the second Ga beam rate among all Ga beam rates during the deposition and growth of the entire second AlGaAs isolation layer, where the number of rotations of the sample holder is an integer, will be used as the third Ga beam rate.

Images