Method for determining molecular beam epitaxy doping process parameters

Abstract

The application provides a method for determining molecular beam epitaxy doping process parameters, and relates to the technical field of semiconductor manufacturing. The method comprises the following steps: growing a Si-doped GaAs epitaxial layer on a GaAs substrate to obtain a first doping curve between a doping concentration and a Si source furnace temperature, and a Ga source furnace power is a first power; fixing the Ga source furnace power as a second power to obtain a second doping curve; converting the second doping curve into a second conversion curve; when growing an epitaxial layer with an expected doping concentration subsequently, obtaining a first Si temperature T1 according to the first doping curve, obtaining a second Si temperature T2 according to the second conversion curve, and calculating T3. By obtaining the doping curve under different Ga source furnace powers, the influence of the Ga source furnace power on the doping concentration is quantified, and when the Si temperature is determined subsequently, the influence of the power on the doping concentration is combined, so that the doping concentration corresponding to the determined Si temperature is closer to the expected doping concentration, and the precision of the doping concentration is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of semiconductor manufacturing, in particular to a method for determining process parameters of molecular beam epitaxy (MBE) doping. BACKGROUND

[0002] For various conventional device structures (for example, pHEMT device structures) based on GaAs (gallium arsenide), when epitaxial wafer growth is performed using MBE (molecular beam epitaxy) technology, process parameters associated with technical indicators such as material thickness composition, doping concentration, etc. need to be calibrated first, and then the epitaxial wafer is grown using the calibrated process parameters.

[0003] For MBE growth of GaAs material, Si is usually used as an N-type doping source. For a given GaAs growth rate, the desired N-type doping concentration can be achieved by selecting a suitable Si source furnace temperature through calibration. In actual production, since the doping concentration of different epitaxial structures is usually different, calibration experiments are usually performed for different doping concentrations, which not only consumes time, but also significantly increases production costs. Therefore, for a specified MBE system and growth rate, multiple doping growths are usually performed by changing the Si source furnace temperature, so as to obtain the corresponding relationship between the Si source furnace temperature and the doping concentration, and within a certain period after obtaining the corresponding relationship, for the epitaxial structure to be grown and the doping concentration, the corresponding Si source furnace temperature is directly determined according to the corresponding relationship.

[0004] The doping concentration is not only related to the Si source furnace temperature, but also related to the actual growth temperature of the substrate. In the MBE system, the sensing temperature of the substrate temperature can be set to the same temperature during growth. The substrate temperature of the conventional MBE system is sensed by a non-contact thermocouple located between the back of the substrate and the substrate heating assembly. Although the thermocouple temperature can reflect the substrate temperature to some extent, due to the blocking of the substrate itself, the thermocouple cannot accurately reflect the temperature change of the substrate surface caused by the front heat source of the substrate (the source furnace as the heat source). Therefore, if the temperature change of the substrate caused by the front heat source of the substrate is different under different growth batches, but the same doping corresponding relationship is still used, the actual doping concentration will deviate from the expected doping concentration, thereby degrading the performance of the epitaxial wafer and the semiconductor device. Therefore, a more accurate method for determining doping process parameters needs to be proposed. SUMMARY

[0005] The present application aims to provide a method for determining process parameters of molecular beam epitaxy (MBE) doping to solve the problem of improving the accuracy of doping concentration in MBE.

[0006] To achieve the above-mentioned purpose, the technical solution adopted by the present application is as follows:

[0007] The application provides a method for determining a molecular beam epitaxy doping process parameter, which comprises the following steps:

[0008] Si-doped GaAs epitaxial layers are grown on a GaAs substrate to obtain doped epitaxial wafers, and doping concentrations are tested, a first doping curve between the doping concentrations and Si source furnace temperatures is obtained by changing the Si source furnace temperature and performing multiple growth tests, in the growth process of obtaining the first doping curve, the power of a Ga source furnace is a first power P1, and the growth rate of Ga is a first rate;

[0009] the power of the Ga source furnace is fixed as a second power P2, P2 is not equal to P1, the growth rate of Ga is a second rate, and multiple growth tests are performed again to obtain a second doping curve between the doping concentrations and the Si source furnace temperatures;

[0010] the second doping curve is converted into a second doping conversion curve corresponding to the first rate according to the first rate and the second rate;

[0011] when GaAs epitaxial layers with an expected doping concentration are grown subsequently, the growth rate of Ga is the first rate, and the Si source furnace temperature is T3, T3 is obtained by the following method: a third power P3 of the Ga source furnace is obtained, a corresponding first Si temperature T1 is obtained according to the first doping curve for the expected doping concentration, a corresponding second Si temperature T2 is obtained according to the second doping conversion curve, and T3=(P3-P1) / (P2-P1)×(T2-T1)+T1 is calculated.

[0012] Optionally, the second doping curve is converted into the second doping conversion curve corresponding to the first rate according to the first rate and the second rate, and specifically comprises the following steps:

[0013] for any Si temperature t in the second doping curve, a corresponding doping concentration N is obtained, a conversion concentration N is calculated according to the following formula: i : N i =N×second rate / first rate, and the second doping conversion curve is constituted by all the temperatures t and the corresponding conversion concentrations N i .

[0014] Optionally, for the Si source furnace temperatures and the corresponding doping concentrations obtained by the multiple growth tests, a corresponding doping curve is obtained by least square polynomial fitting.

[0015] Optionally, the thicknesses of the Si-doped GaAs epitaxial layers grown to obtain the first doping curve and the second doping curve are equal.

[0016] The application has the following beneficial effects:

[0017] The method for determining a molecular beam epitaxy doping process parameter provided by the application comprises: growing a Si-doped GaAs epitaxial layer on a GaAs substrate to obtain a doped epitaxial wafer, testing the obtained doping concentration, changing the Si source furnace temperature and performing multiple growth tests to obtain a first doping curve between the doping concentration and the Si source furnace temperature, in the growth process of obtaining the first doping curve, the power of the Ga source furnace is a first power P1, and the growth rate of Ga is a first rate; fixing the power of the Ga source furnace as a second power P2, and P2 is not equal to P1, the growth rate of Ga is a second rate, and multiple growth tests are performed again to obtain a second doping curve between the doping concentration and the Si source furnace temperature; according to the first rate and the second rate, the second doping curve is converted into a second doping conversion curve corresponding to the first rate; in subsequent growth of a GaAs epitaxial layer with an expected doping concentration, the growth rate of Ga is the first rate, and the Si source furnace temperature is T3, T3 is obtained by the following method: obtaining a third power P3 of the Ga source furnace, according to the first doping curve, obtaining a corresponding first Si temperature T1 for the expected doping concentration, according to the second doping conversion curve, obtaining a corresponding second Si temperature T2, and calculating T3=(P3-P1) / (P2-P1) x(T2-T1)+T1. By obtaining the first doping curve and the second doping curve under different Ga source furnace powers, the influence of the Ga source furnace power on the doping concentration is quantified, and the influence of the power on the doping concentration is considered in subsequent determination of the Si doping temperature for epitaxial growth, so that the doping concentration corresponding to the determined Si doping temperature can be closer to the expected doping concentration, and the accuracy of the doping concentration is improved. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative effort.

[0019] Figure 1 A flowchart of the method for determining a molecular beam epitaxy doping process parameter provided by the embodiments of the present application is shown. DETAILED DESCRIPTION

[0020] The technical solutions in the embodiments of the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present application.

[0021] For the molecular beam epitaxy growth of GaAs material, the doping concentration of N-type doping is not only related to the Si source furnace temperature, but also related to the actual growth temperature of the substrate. The substrate temperature of the conventional molecular beam epitaxy system is sensed by a non-contact thermocouple located between the back of the substrate and the substrate heating assembly. Although the thermocouple temperature can reflect the substrate temperature to some extent, the thermocouple cannot accurately reflect the temperature change of the substrate surface caused by the heat source (source furnace as the heat source) on the front of the substrate due to the blocking of the substrate itself. Therefore, if the temperature change of the substrate caused by the heat source on the front of the substrate is different in different growth batches, but the same doping relationship is still used, the actual doping concentration will deviate from the expected doping concentration, and the performance of the epitaxial wafer and the semiconductor device will be degraded. Therefore, a more accurate method for determining the doping process doping needs to be proposed.

[0022] In the actual molecular beam epitaxy growth process, in addition to the substrate heating assembly of the molecular beam epitaxy system which can heat the substrate temperature, the various source furnaces used will release a certain amount of heat to the substrate surface (the front of the substrate) through thermal radiation after the shutter of the source furnace is opened, thereby also affecting the temperature of the front of the substrate to some extent. Since the source furnace is located on the front of the substrate and the thermocouple is located on the back of the substrate and is non-contact with the substrate, the temperature rise of the front of the substrate caused by the heat provided by the source furnace cannot be accurately reflected by the thermocouple, thereby causing the actual temperature of the substrate surface to change when the heat provided by the source furnace changes, even if the substrate temperature set in the molecular beam epitaxy system does not change (that is, the thermocouple sensing temperature does not change), the actual temperature of the substrate surface will change, and this change will cause the Si doping concentration to change. In the case where the Si source furnace temperature does not change (that is, the dopant dose does not change), the higher the substrate surface temperature, the higher the doping efficiency, and thus the higher the doping concentration. For the molecular beam epitaxy growth of Si-doped GaAs epitaxial layer, considering the actual situation of the source furnace structure, the additional heat brought to the substrate surface is mainly provided by the Ga source furnace when the shutters of the Si source furnace, the Ga source furnace and the As source furnace are all opened. Therefore, in the optimization technical solution of the present application, only the influence of the heat provided by the Ga source furnace on the doping concentration is considered.

[0023] Figure 1 A flowchart of a method for determining molecular beam epitaxy doping process parameters provided by an embodiment of the present application is shown. As shown in Figure 1 The method for determining molecular beam epitaxy doping process parameters provided by the present application includes:

[0024] Step 101, growing a Si-doped GaAs epitaxial layer on a GaAs substrate to obtain a doped epitaxial wafer, testing to obtain a doping concentration, through changing the Si source furnace temperature and performing multiple growth tests, obtaining a first doping curve between the doping concentration and the Si source furnace temperature, in the growth process of obtaining the first doping curve, the power of the Ga source furnace is a first power P1, and the growth rate of Ga is a first rate.

[0025] Since the heat delivered by the source furnace to the substrate comes from the heating assembly of the source furnace, the higher the heating power of the source furnace, the more heat can be delivered to the substrate when the source furnace shutter is opened. Therefore, the amount of heat delivered by the source furnace to the substrate can be characterized by obtaining the heating power of the source furnace. The power of the Ga source furnace can be obtained by the molecular beam epitaxy system itself.

[0026] For the conventional growth of GaAs epitaxial layer, the process parameter to be controlled is the growth rate of Ga, and the higher the growth rate of Ga, the higher the power of the Ga source furnace in general. In addition, from a long-term (e.g., several days or more) point of view, as the Ga source material is used, the remaining Ga source material in the Ga source furnace gradually decreases, and in order to maintain the same growth rate, the power of the Ga source furnace slowly increases. For a short period (e.g., several or tens of hours), the power of the Ga source furnace can be considered to be stable and unchanged at a constant rate. However, for different rates or different periods, the power of the Ga source furnace can be variable. For the growth rate of Ga being a first rate, multiple epitaxial growths (changing the Si source furnace temperature) can be continuously performed in a short period, and the Ga source furnace powers corresponding to these multiple epitaxial growths are a first power P1. Therefore, the different doping concentrations obtained by these multiple growths can be considered to be related only to the Si temperature, and it should be understood that the substrate thermocouple temperatures of these multiple epitaxial growths are the same, and the thicknesses of the grown GaAs epitaxial layers are also the same. After obtaining multiple epitaxial wafer samples corresponding to the Si source furnace temperature one by one, the doping concentration tests are respectively performed, and through data fitting, a fitting curve between the doping concentration and the Si source furnace temperature can be obtained, and the fitting curve is taken as the first doping curve.

[0027] Step 102, fixing the power of the Ga source furnace to a second power P2, and P2 is not equal to P1, the growth rate of Ga is a second rate, and multiple growth tests are performed again to obtain a second doping curve between the doping concentration and the Si source furnace temperature.

[0028] The epitaxial growth in step 102 can be performed immediately after the epitaxial growth in step 101 is completed. In actual operation, in order to change the power of the Ga source furnace, specifically, the set temperature of the Ga source furnace for growth can be changed, and then the corresponding power is obtained from the molecular beam epitaxy system, and the corresponding growth rate of Ga can be obtained through rate calibration experiments or rate empirical curves. Again, multiple growth tests are performed, and the doping concentration is tested and data is fitted to obtain a second doping curve between the doping concentration and the Si source furnace temperature.

[0029] Generally, the doping concentration has no absolute relationship with the thickness of the epitaxial layer, but in order to reduce measurement error, a relatively thick Si-doped GaAs epitaxial layer can be epitaxially grown, and the thickness of the Si-doped GaAs epitaxial layer grown for obtaining the first doping curve and the second doping curve is equal.

[0030] Optionally, the Si source furnace temperature and the corresponding doping concentration obtained through the multiple growth tests in steps 101 and 102 can be fitted into a corresponding doping curve through least squares polynomial fitting.

[0031] Step 103: converting the second doping curve into a second doping conversion curve corresponding to the first rate according to the first rate and the second rate.

[0032] The doping concentration in steps 101 and 102 can be obtained through Hall effect measurement or electrochemical measurement. It should be understood that the doping concentration in steps 101 and 102 is for the corresponding growth rate. For example, for the same Si dopant dose (the dopant dose corresponds to the Si source furnace temperature), if the growth rate of the GaAs epitaxial layer is higher, the obtained doping concentration is lower. Therefore, in order to make the doping curves obtained for different growth rates comparable, the second doping curve obtained for the second rate can be converted into a second doping conversion curve corresponding to the first rate, and the second doping conversion curve obtained thereby can be considered as a doping curve obtained at the first rate. Specifically, the conversion of the second doping curve into the second doping conversion curve corresponding to the first rate according to the first rate and the second rate specifically includes: for any Si temperature t in the second doping curve, obtaining the corresponding doping concentration N (that is, obtaining the doping concentration N corresponding to the temperature t in the second doping curve), calculating the conversion concentration N i : N i =N×second rate / first rate, and the second doping conversion curve is composed of all temperatures t and the corresponding conversion concentrations N i In actual operation of obtaining the second doping conversion curve, N iThe function relationship of the second doping conversion curve is obtained by transforming the function relationship of the second doping curve directly in relation to N.

[0033] In step 104, when growing the GaAs epitaxial layer of the expected doping concentration subsequently, the Ga growth rate is the first rate, the Si source furnace temperature is T3, T3 is obtained by the following way: obtaining the third power P3 of the Ga source furnace, obtaining the corresponding first Si temperature T1 according to the first doping curve for the expected doping concentration, obtaining the corresponding second Si temperature T2 according to the second doping conversion curve, and calculating T3=(P3-P1) / (P2-P1)×(T2-T1)+T1.

[0034] After obtaining the first doping curve corresponding to P1 and the second doping conversion curve corresponding to P2 (both the first doping curve and the second doping conversion curve are curves corresponding to the first rate), the GaAs epitaxial layer with the expected doping concentration is grown at the first rate, at this time the power of the Ga source furnace corresponding to the first rate of the Ga growth rate is the third power P3, through linear interpolation, the corresponding Si source furnace temperature at this time can be calculated as T3=(P3-P1) / (P2-P1)×(T2-T1)+T1. In particular, if P3=P1, T3=T1, at this time the corresponding Si temperature can be directly determined according to the first doping curve, and if P3=P2, T3=T2, at this time the corresponding Si temperature can be directly determined according to the second doping conversion curve. It should be understood that if the Ga growth rate is not equal to the first rate (for example, the Ga growth rate is equal to the third rate) when the GaAs epitaxial layer with the expected doping concentration is grown subsequently, the initial expected doping concentration corresponding to the third rate can be converted to the third doping concentration corresponding to the first rate according to the conversion method described in step 103, and then the third doping concentration is taken as the expected doping concentration described in step 104, and the calculation is performed in the manner of step 104. For the expression of T3, if P2>P1, since the second doping conversion curve corresponds to a higher doping efficiency for the same doping concentration, T2<T1, further, if P3>P1, the value of (P3-P1) / (P2-P1)×(T2-T1) is negative, so T3<T1. The expression of T3 quantifies the influence of the power of the Ga source furnace on the doping concentration by using linear interpolation. In order to make the calculation result more accurate, P2>P1, and the value range of P3 is: P1<P3<P2. In order to make the value of P3 between P1 and P2, when obtaining the second doping curve, a larger P2 can be selected compared to P1, and the corresponding second rate is also larger, when growing the GaAs epitaxial layer with the expected doping concentration subsequently, since the Ga growth rate is the first rate, it is easy to obtain P3<P2, and at this time the remaining amount of source material in the Ga source furnace is less than the remaining amount when obtaining the first doping curve, at this time in order to maintain the same first rate, the power P3 of the corresponding Ga source furnace must be greater than the previous power P1.

[0035] It should be noted that in the whole process of growing the GaAs epitaxial layer (including the growth of the GaAs epitaxial layer for obtaining the first doping curve, the growth of the GaAs epitaxial layer for obtaining the second doping curve, and the subsequent growth of the GaAs epitaxial layer with the expected doping concentration), the Ga source furnace for providing the Ga molecular beam source is the same source furnace. Here, the "same source furnace" only means that the number of Ga source furnaces for providing the Ga molecular beam source is 1, and they are located at the same position on the molecular beam epitaxy system. Since the source material is consumed during the molecular beam epitaxy growth, the remaining source material in the source furnace is slowly and gradually reduced. Here, the "same source furnace" does not limit the amount of Ga source material remaining in the source furnace.

[0036] In summary, by obtaining the first doping curve and the second doping curve under different Ga source furnace powers, the influence of the Ga source furnace power on the doping concentration is quantified, and the influence of the power on the doping concentration is considered when the Si doping temperature for epitaxial growth is subsequently determined, so that the doping concentration corresponding to the determined Si doping temperature can be closer to the expected doping concentration, and the accuracy of the doping concentration is improved.

[0037] The above embodiments are only for illustrating the technical concepts and characteristics of the present application, and the purpose is to enable those skilled in the art to understand the content of the present application and implement it, and cannot limit the protection scope of the present application. Any equivalent changes or modifications made in accordance with the spirit and essence of the present application should be covered within the protection scope of the present application.

Claims

1. A method for determining molecular beam epitaxy doping process parameters, characterized in that, The method includes: A Si-doped GaAs epitaxial layer is grown on a GaAs substrate to obtain a doped epitaxial wafer. The doping concentration is measured. By changing the Si source furnace temperature and performing multiple growth tests, a first doping curve is obtained between the doping concentration and the Si source furnace temperature. During the growth process of obtaining the first doping curve, the power of the Ga source furnace is the first power P1, and the growth rate of Ga is the first rate. The power of the Ga source furnace is fixed at the second power P2, and P2 is not equal to P1. The growth rate of Ga is the second rate. Multiple growth tests are performed again to obtain the second doping curve between the doping concentration and the temperature of the Si source furnace. Based on the first rate and the second rate, the second doping curve is converted into a second doping conversion curve corresponding to the first rate; During the subsequent growth of GaAs epitaxial layers with the expected doping concentration, the Ga growth rate is the first rate, and the Si source furnace temperature is T3. T3 is obtained as follows: the third power P3 of the Ga source furnace is obtained; for the expected doping concentration, the corresponding first Si temperature T1 is obtained according to the first doping curve; the corresponding second Si temperature T2 is obtained according to the second doping conversion curve; and T3 is calculated as T3 = (P3 - P1) / (P2 - P1) × (T2 - T1) + T1. The step of converting the second doping curve into a second doping conversion curve corresponding to the first rate based on the first rate and the second rate specifically includes: For any Si temperature t in the second doping curve, obtain the corresponding doping concentration N, and calculate the conversion concentration N according to the following formula. i :N i =N × second rate / first rate, and derived from the total temperature t and the corresponding conversion concentration N i This forms the second doping conversion curve.

2. The method for determining molecular beam epitaxial doping process parameters according to claim 1, characterized in that, The Si source furnace temperature and corresponding doping concentration were obtained from multiple growth tests, and the corresponding doping curves were obtained by least squares polynomial fitting.

3. The method for determining molecular beam epitaxial doping process parameters according to claim 1, characterized in that, The thickness of the Si-doped GaAs epitaxial layer grown by the first doping curve and the second doping curve is the same.

Images