JPH0573326B2 - - Google Patents

Description

[Detailed description of the invention]

[Industrial Application Field] The present invention relates to a semiconductor vapor phase growth apparatus, and in particular, in a plurality of sequence processes for performing one vapor phase growth, process parameters required for each sequence process are determined. A process program that includes a set of information regarding time, gas, and temperature is formed as a unit, and an operator can execute this process program as a unit while modifying it interactively.It is capable of general-purpose automatic control. The present invention relates to a semiconductor vapor phase growth apparatus. [Prior Art] Today, as a method for manufacturing semiconductor chips, vapor phase growth equipment that performs vapor phase growth on semiconductor wafers is increasingly used, and automation of the operation of this type of equipment is also required. It became. Currently used vapor phase growth apparatuses generally use a method in which a sequence program (hereinafter referred to as a process program) that instructs the progress of a process in a reactor is specified using a pinboard switch or the like. There is. In this case, regarding the specification of the flow rate of the gas used and the temperature inside the furnace,
The operator adjusts and sets the variable resistor attached to the control device, and the control requires a lot of skill and experience. For example, FIG. 1 shows a system configured to control the progress of a process within a reactor using a pinboard switch system. In FIG. 1, pins are inserted on the setting panel A of the pinboard switch in the order in which they are executed for each process program PPi (i=1 to 17), and the sequence of process programs in each designated order is inserted. Time is hours, minutes,
It can be set in units of seconds. Also,
Relay ladder circuit B has process programs PP2, PP3, PP1, PP4, PP6, corresponding to the sequence order, ie, steps,...
Commands to enable the contents of PP5 and PP7 are given in this order, and control signals are given to valve devices and the like corresponding to the processes of each command. However, with this pinboard switch control method, only the time for each process sequence can be set, and the gas flow rate and furnace temperature used in that process must be set using another control object (variable resistor). Must be. In addition to the pinboard switch method shown in Figure 1, control methods using general-purpose sequence controllers have also been implemented, but in these control methods, the flow rate and temperature are directly specified using programmed data. It is not equipped with a setting device such as a variable resistor. From this point of view, in order to reduce the operator's work as much as possible, control systems have been proposed in which process parameters such as gas flow rate, temperature, and time are controlled by a computer using a DDC method. [Problem to be solved by the invention] Today, the manufacturing process for semiconductors, from raw material to completion as a chip, requires many more steps than other industries, and the vapor phase growth process is just one of them. .
Various processes are organically combined in a relationship that leads to completion as a chip, and it is difficult to significantly increase the number of devices that handle one single process part from the standpoint of manufacturing process and line balance. be. In addition, due to today's unbearable demands for increasing the degree of integration of semiconductor chips and demands for improving accuracy, in the technological development of semiconductor manufacturing equipment, for example, elucidation of physical and chemical phenomena in CVD reactors,
The development of new measurement methods is progressing with complex effects. In particular, in the case of a vapor phase growth apparatus, work such as correction and correction of data regarding film thickness, etc. is required while the apparatus is in operation. A specific example is as follows. Changes in the liquid level in the bubbler tank change the growth rate of the film thickness, and the operator must perform periodic inspections and respond by changing or modifying the source gas flow rate or sequence time. If a uniform temperature distribution is not guaranteed over the entire surface of the susceptor used to place the wafer, the film thickness distribution and resistivity distribution will deteriorate, but this may be due to variations in the manufacturing process of the susceptor. Uniform heat adjustment must be performed. In particular, when replacing a new susceptor and increasing the temperature of the susceptor using an existing process program, it is necessary to remove gas and adjust the temperature. A uniform heat distribution can be obtained. In addition, even if the uniform heat distribution is good, it does not necessarily mean that a good product will be obtained, so a monitor wafer (evaluation wafer) is introduced and the film is applied once, and if there is a large variation in the results, the film is applied again. Repeat the cycle of soaking, adjusting and raising the temperature. If the film thickness and resistivity specifications for the wafer are changed during the vapor phase growth process, two or three monitor wafers are placed on the susceptor and a test process is executed to ensure that the specifications are met. This requires evaluation work by a process engineer to determine this while varying the conditions, and data correction work by an operator. Generating a vapor phase grown film on a wafer is
A grown film is also produced on the susceptor, and since this produced film has an adverse effect on the wafer (autodoping), it is necessary to remove this produced film layer. In this case, the operator must monitor the removal status of the film on the susceptor in the reactor, and if the removal is not in good condition, it is necessary to temporarily stop the control (sequence hold) and continue the film removal work. Note that the life of the susceptor is significantly affected by the heating time and the number of cycles, so the susceptor must be replaced periodically. In this case, the new susceptor needs to gradually increase its power while measuring the temperature in front of the reactor to reduce the temperature shock, and for this reason the predetermined parameters during the heating process can be adjusted on site by the operator. Correction work is also required. However, with conventional control systems that use pinboard switch systems or sequence controllers, as mentioned above, when manufacturing chips of a certain quality, there are many aspects to the operation of semiconductor vapor phase growth equipment, such as gas flow rates and temperature conditions. Because it is necessary to repeatedly change the condition settings, even if the order of the process program and its sequence time can be set programmably, the gas flow rate and furnace temperature cannot be changed by the operator using a separate setting device (variable resistor, etc.). It is designed to be operated directly. From an operator's perspective, this has the advantage of being able to operate the reactor with peace of mind; however, it is not possible to record when appropriate conditions are obtained or immediately determine the relationship between these and the sequence contents, so the equipment During operation, operator judgment factors account for most of the control operations. In addition, gas flow rate, temperature, time, etc. can be set as process parameters and
In a control system using the DDC method, it is necessary to modify the contents of the process program in order to change the condition settings as described above. You have to build a program. however,
In this type of control system, once execution of a preset process program is started, the contents of the process program cannot be modified until the operation is stopped midway or one processing operation is completed. For this reason, it is not possible to easily change the condition settings as described above, which has the drawback of not only lowering the yield of products but also lowering the operating rate of the apparatus. Furthermore, this type of semiconductor vapor phase growth apparatus is usually equipped with a plurality of reactors, and productivity can be improved by operating each of these reactors in a sequential cycle. . However, in this case, for one cycle of operation of each reactor, the heating process that maintains the temperature inside the furnace at a predetermined temperature requires a lot of processing time to adjust the gas atmosphere inside the furnace before and after the heating. In short, especially when each reactor is controlled by a series of process programs, there is a drawback that it takes a long time until the next reactor starts operating. SUMMARY OF THE INVENTION Therefore, an object of the present invention is to provide process parameters such as time, gas, and
Temperature-related information is stored as a process program, and an operator calls up the contents of this process program as a unit and displays it on a display means, making it possible to modify each process parameter and to execute the contents in real time. When a plurality of reactors are sequentially operated in a cycle, the switching timing of the heating process is set continuously so that the heating of the second reactor can be started at the same time as the heating of the first reactor is completed, and therefore, the switching timing of the heating process is Before and after the switching timing, the first reactor and the second reactor each execute a series of process programs in parallel regarding gas atmosphere adjustment in the reactor, thereby efficiently realizing cycle operation of multiple reactors. An object of the present invention is to provide a semiconductor vapor phase growth apparatus that can perform semiconductor vapor phase growth. [Means for Solving the Problems] A semiconductor vapor phase growth apparatus according to the present invention includes a plurality of reactors for performing vapor phase growth on a substrate such as silicon, a means for heating each of the substrates, A network of pipes connecting various gas sources necessary for vapor phase growth and each of the reactors, and a network of pipes provided on the network of pipes to guide desired amounts of the various gases to each of the reactors. A semiconductor vapor phase growth apparatus comprising a valve device for forming the pipe network, a control device for providing a signal for controlling on/off or the degree of opening of the valve device, and a signal for controlling each of the heating means, A heating means is arranged for each substrate placed in each reactor, and a changeover switch selectively connects each of the heating means to a common power source for supplying heating power, and a start switch that starts each of the reactors, respectively. and further comprising: (a) a process program group to be executed in each reactor; (c) the time from the start of the process to the heating start command of the second process program group corresponding to the second reactor following the first reactor; , and (d) from the time when the second start switch corresponding to the second reactor is activated, the heating power to the second reactor is increased after the heating power supply to the first reactor is completed. The method is characterized in that a calculation unit is provided for setting a process sequence start time of the second reactor so that supply is started. [Function] According to the semiconductor vapor phase growth apparatus according to the present invention, when performing heating and gas supply in the reactor,
switching on of the valve arrangement in which the gas supply is installed in the pipe network;
The heating means is controlled by turning off or its opening degree, and the heating means is controlled by a predetermined signal.
Information regarding the required sequence time, the type and flow rate of gas to be supplied during this sequence time, and the temperature to be achieved in the furnace in accordance with each actual sequence process related to vapor phase growth. A group of process programs is constructed by storing them as data and forming process programs, and each process program can be called at will and modified interactively, and it can be immediately executed in a running process. The system is configured so that operations can be performed smoothly in response to changes in each sequence process in actual physical and chemical processes. In particular, in the apparatus of the present invention, when a plurality of reactors are sequentially operated in cycles, the first reactor is By configuring a part of the process program group for the second reactor and the process program group for the second reactor to be executed in parallel, the heating process is sequentially and continuously cycled for each reactor. This makes it possible to efficiently utilize heating power and shorten cycle time, thereby improving the operating rate of the device. [Example] Next, an example of the semiconductor vapor phase growth apparatus according to the present invention will be described with reference to the accompanying drawings starting from FIG. FIG. 2 is an external view showing an embodiment of the semiconductor vapor phase growth apparatus of the present invention. In FIG. 2, reference numeral 11 is a high frequency generator, and 12 and 13 are reactor R.
1. This is the main body of a vapor phase growth apparatus equipped with R2. Reference numeral 14 is a control section, which controls the gas flow rate into each reactor, the temperature inside each reactor, etc., and 14 is an operation panel of the control section 14, which includes an operation key input section,
Including display units, etc. The details are in the 5th section.
As shown in the figure. Reference numbers 12A and 13A are reactor R
1. This is an operation panel for opening and closing R2. FIG. 3 is a conceptual block diagram illustrating the operations performed in the vapor phase growth apparatus shown in FIG. A process program previously stored in an internal memory 17 provided in the unit 14 is provided to the processing unit 14-1, and in this case, the device drive unit 1 is processed in response to the input process program.
Opening/closing of the valve device etc. provided at 6 and the operation for controlling the opening degree thereof are performed. Further, the contents of the processing of the processing section 14-1 and the input information to the processing section 14-1 are displayed on the display 14A-1.
It is now possible to display on 2. In the apparatus bodies 12 and 13 corresponding to FIG.
1 and R2, the gas set by the device drive section 16 is configured to flow into the gas. Note that the above-mentioned equipment drive unit 16, necessary pipes, valves, etc.
It is disposed at the bottom of the back side of each device 12, 13 in the same way as shown in the figure. Further, FIG. 7 shows details of the pipe network showing the layout of pipes and valves. FIG. 4 is a detailed block diagram of the control system of the apparatus according to the present invention. In the figure, reference numeral 21 is a central processing unit CPU of the main computer, and this CPU 21 includes a data bus 22, an i/
o bus 23 is connected. The data bus 22 includes a storage unit 24 that stores in advance a series of process programs to be executed in each reaction path R1 or R2, a CRT that temporarily stores the contents to be displayed on a display device CRT 29, and a RAM 2.
5. Furthermore, a temporary storage section 26 and a storage section 27 storing various processing programs for operating this system are connected, respectively. The above-mentioned temporary storage unit 26 stores data used while the system is in operation, such as input data from the keyboard 31 and various switches.
It is used to store ON/OFF information or a group of process programs provided from an external storage medium such as a cassette tape. Furthermore, a CRT interface 28 is connected to the I/O bus 23, and the content to be displayed on the CRT 29 is transferred to the CRT 29.
Give to 9. Further, reference numerals 30 and 32 are input modules, which act to input data from the keyboard 31 and signals from switches such as the pressure switch PS and the limit switch LS into the temporary storage section 26, respectively. Furthermore, the output module 3 is connected to the I/O bus 23.
4, 36, and 38 are connected to output parts 35 such as lamps, LEDs, and valves, gas valves 37,
An output command is given to the relay drive unit 39. Reference numeral 40 denotes a motor and a valve driven by the relay drive section 39, which respectively react R1 and R1.
These are the motor for rotating the susceptor in R2 and the valve for driving the cylinder for opening and closing the furnace lid. Furthermore, a D/A conversion module (DAM) 41 and an A/D conversion module (ADM) 43 are connected to the I/O bus 23, and the DAM 41 controls the flow rate of gas flowing through the flow control valves MFC TiC and VC1. Give the specified control voltage as an analog quantity. Further, the ADM 43 receives an analog signal from a detection section that detects the flow rate flowing to each control valve as a feedback signal, and converts this into a digital signal. 46
is a sub-central processing unit, and is used when transferring a process program group stored on a cassette magnetic tape (CMT) 51 to a temporary storage unit (RAM) 53 via an interface 49 for storage, or when transferring it to a temporary storage unit (RAM) 53 for storage. On the other hand, it operates to write the process program group stored in the RAM 53 thereto. This CPU 46 is a storage ROM
It operates according to the processing program stored in 52. 48 is the ROM52 and RAM5 mentioned above.
3 is a data bus that connects the CPU 46, and 54 is an I/O bus that connects the interface 49 and a high speed memory data transfer unit (HMT) 50 with the CPU 46. On the other hand, another high-speed memory data transfer unit (HMT) 45 is arranged on the aforementioned I/O bus 23, and this HMT 45 and HMT 50 are connected by a data highway 47, so that
Transfer the contents of RAM53 to RAM26 or vice versa
It acts to transfer the contents of RAM 26 to RAM 53 at high speed. By doing this,
It is possible to avoid the problem that the time required to read a process program from the CMT 51 (or a magnetic card, etc.) and write the same data to the CMT 51 limits the speed of the arithmetic processing of the CPU 21. Of course, instead of elements 46-55, these can be used as i/
Data may be exchanged via an input/output module connected to the o bus 23. In addition, the types of processing programs shown in the ROM 27 include sequential reading of a process program group (hereinafter referred to as PPG) stored in the RAM 26;
The following processing program is used so that the CPU 21 decodes this into sequence instructions corresponding to each process program (hereinafter referred to as PP). A processing program for controlling the CPU 21, that is, a process program processing program (PROCESS・C) A modification processing program (MODiFY) that controls the CPU 21 to modify the contents of the PPG stored in the RAM 26, which uses the keyboard 31 to read the necessary data A process program generation processing program (PROCESS) for inputting and generating a new PPG, a RUN processing program for displaying the currently ongoing process on the CRT29, and an arbitrary PP(i) in the PPG. Processing program (STEP) that converts to PP(j), RAM to external storage medium (e.g. CMT51)
A processing program (STORE) for transferring the PPG stored in the RAM 26 via the RAM 53, a processing program (SORT) for performing the opposite operation to the store function, a processing program PROCESS・C for transferring the PPG stored in the RAM 26. A verification processing program (VERiFY) to confirm this before applying the system, a processing program (DiAGNOSiS) that performs self-diagnosis during operation of this system, a process for servicing the elapsed time during operation of one PPG, elapsed time A station program (USED TiME) that calculates , a processing program (TEST) that performs various test functions, etc. are stored in the ROM 27 as processing programs, and by specifying one of these,
The CPU 21 is designed to perform necessary arithmetic functions according to each processing program. The details of the operation of each processing program in the ROM 27 shown in FIG. 4 will be explained with reference to the flowchart described later. 5 is a front view of the panel operation panel of the control device shown in FIG. 2. FIG. In the figure, reference numeral 61
62 is a display device, 62 is a cassette tape mounting section for mounting a cassette magnetic tape, 63 is a key input device, 64 is a temperature control section, 64-1 is a furnace temperature, and 64-2 is a temperature setting switch.
Further, 65 is a display section that displays the type of process program group PPG. In area 66,
Alarm buzzer 66-1 and alarm reset button 6
6-2 is provided. Data input area 67
There are a program start push button 67-1, a gas selection switch 67-2, a reactor selection pattern thumbwheel switch 67-3, and a PPG selection designation thumbwheel switch 67-4. Switch 67-3 is 0→R1 only, 1→R2 only, 3→R
It is like 1+R2. Reference numeral 67-5 is a push button, which becomes effective when both push buttons 67-1 and 67-1 are pushed. In the area 68, 68-1 indicates a start switch, and 68-2 a stop switch, which is used as a command to start the sequence process of one PPG. Furthermore, 6
8-3 is an induction heating furnace, and 69 corresponds to the type of process in which the reaction is progressing in the upper row.
It is designed to turn on the LED. Each PPG consists of a suitable combination of 1 to 17 sequential processes. In the lower row are various alarm displays.
Equipped with LED. FIG. 6 is a detailed cross-sectional view of reactor R1 or R2. In the figure, an inlet 72 for gas used for vapor phase growth in the furnace is provided at the lower center of the bottom plate 71, and the gas rises in a conduit 74 extending upward from the center of the bottom plate 71 and reaches the top. Exhaust vent 7
It is designed to be discharged from 3 onwards. Furthermore, a rotary member 76 is disposed on the outer circumference of the above-mentioned conduit 74, and the rotary member 76 supports the susceptor 75 at its upper part and the rotary member 76 at its top.
It is designed to be rotated by 7. An induction heating coil 79 is arranged below the susceptor 75 with a cover 78 in between. Further, 80 is an insulating plate that also serves as a weight support for the coil 79, and is fixed above the bottom plate 71 with bolts 81. Further, 82 and 83 are joint portions for connecting the induction heating coil 79 to the outside. Water is allowed to flow inside the coil to prevent the heat generated by the high-frequency current from damaging the coil itself. The ceiling cover 90 facing the bottom plate 71 is made up of three layers.
Each of them consists of a quartz layer 84, a first stainless steel layer 85, and a second stainless steel layer 86 from the inside. There are gaps between each layer 84, 85, 86. Also, 8
Reference numeral 8 denotes a clamp member which presses the flange 87 of the ceiling cover 90 downward when the air cinder device 89 is excited. An observation window 92 for observing the susceptor 75 and the wafer 91 on the susceptor 75 is attached to the ceiling cover 90. Further, a temperature detection window 93 is provided on the ceiling cover 90 to which a sensor TS is attached to detect the temperature of the wafer 91 and the susceptor 75 using light entering through the quartz layer 84. Also,
Reference numeral 94 denotes a bracket integrally constructed with the ceiling cover 90, and is connected to a piston of a cylinder (not shown) so that it can move up and down. For example, when replacing the wafer 91 of the susceptor 75, the ceiling cover 90 is moved upward. FIG. 7 is a gas piping system diagram connected to reactors R1 and R2. In the same figure, reactor R
The gases supplied to 1 and R2 are N ₂ , H ₂ , D _N , D _P and HCl gas chambers 101 , 102 , 103 , 104 , 1 from the left in the lower part of the figure.
05. Further, 106 is a bubbling chamber containing a liquid of silicon tetrachloride SiCl ₄ or SiHCl ₃ . A pressure switch PS1 and a normally open valve 1 (hereinafter, a normally open valve is marked with a minus sign) are provided in a conduit extending upward from the chamber 101, and communicates with a valve PV7. Similarly, a pressure switch PS2 and a valve PV2 are provided in the pipe line extending upward from the chamber 102.
Leads to valve PV8. The outlet ports of valves PV7 and PV8 merge to form mass flow valve MFC1, respectively.
It is connected to PL1 and PL2 via MFC2. Further, on the pipe PL1, there is a gas merging valve PV19,
PV20 is provided between the reaction path R1,
Valve the gas supplied by pipes PL1A and PL2A
Mixing is possible by exciting PV19 and PV20. Similarly, on the pipe line PL2, there is a gas merging valve PV21,
PV22 is provided between the reactor R2,
Valve the gas supplied through pipes PL1A and PL2A.
Mixing is possible by exciting PL21 and PL22. From the chamber 106, two pipes PL3A,
PL3B is extended and connected to valve VC1. Gas _H2 enters port PO of the same valve VC1,
_H2 exits port P2 and connects to pipe PL3A and valve PV3.
It enters the bubbling chamber 106 through A and is discharged into liquid SiCl ₄ to perform bubbling. Therefore, a gas mixture of vaporized SiCl ₄ and H ₂ is created in the space inside the chamber 106, and this gas flows through the pipe.
It is connected to the conduit PL6A through the input port P3 and output port P1 of the valve VC1 through the valve PV3B on PL3B. Dopant N gas chamber 1
Mass flow valves MFC4, MFC5, and MFC6 are connected to the conduit extending upward from 03 via valve PV5. The input side of the valve MFC4-6 is connected to a pipe line PL5 for hydrogen gas _H2 via a valve PV9, and the mixed gas is connected to the pipe line PL1A from the output side so that it can be supplied. A similar piping system is used for dopant P gas chamber 1.
It is also formed for a conduit extending upward from 04. Each valve PV23, PV24, MFC8, MFC
9. It will be understood from the drawings that the MFC 10 corresponds to each of the above-mentioned valves. A valve PV6A and a mass flow valve MFC7 are provided on the pipe line extending from the gas HCl chamber 105, and the merging valves PV20 and PV22 are input to the mixing port, respectively. The mass flow valve MFC7
A pipe line PL5 is joined to the upstream side of the pipe via a valve 6. FIG. 8 shows how a command value voltage is given to the mass flow 42 from the control device including the MASTER CPU 21 shown in FIG. 3, and the command value voltage is given via each D/A converter 123. Outputs from the flow rate detectors attached to each mass flow are sequentially taken out by an analog multiplexer 122, and taken in as digital signals by an A/D converter 121 to the aforementioned control device. In Figure 9, the numbers 1 to 17 in the left column correspond to each process program, and the TiME
The column shows the duration of the sequence in minutes and seconds. Also, the flow rate of the gas used in the sequence is now written on the right side of the TiME column.
A GAS FLOW column is provided. GAS FLOW
A temperature setting column for specifying the temperature θ° C. inside the reactor is provided on the right side of the column. Now, a case where the operation table shown in FIG. 9 is applied to the pipe line shown in FIG. 7 will be described. In FIG. 9, the contents of process program PP1 are N ₂ Purge
and supply N ₂ for 3 minutes at a flow rate of FN1/min. From the chamber 101 in _FIG .
passes through valve 1, valve PV7, MFC1, and then PV
19, enters the reactor P1 through the PV 20 and reaches its exhaust port where parsing is performed. Furthermore, the flow rate
FN1/min is given as a voltage command value to MFC1. Process program PP2 is
H ₂ Purge, a flow rate of FH2/min is set for 3 minutes. Gas _H2 is PV2, PV8, MFC
1, enters the reactor R1 through PV19 and PV20, and is discharged in the same way as in the case of N ₂ Purge. The next process program PP3 (hereinafter referred to as PPi) is HEAT ON1, the amount of gas _H2 supplied to the reactor R1 is FH2/min, and the state of each valve is not changed. Then, the induction heating furnace is set to the first stage level and heated for 3 minutes to reach the first set temperature θ1. Next, in PP4, the gas H ₂ is heated at the second stage level so that the set temperature θ2 is maintained at the same flow rate for 3 minutes. Next, PP5 is HCl VENT. The settings are 3 minutes, _H2 is FH/min, HCl is
The flow rate is FHCL/min. Gas HCl Valve PV
It flows through 6A and MFC7 to Ventoro. HCl
The flow rate of FHCL1/min is set by the command voltage to the MFC7. Next, PP6 is HCl ETCH and lasts for 3 minutes. Therefore, for PP5, the merge valve PV20
It joins with H ₂ and is supplied to reactor R1. Next, H ₂ Purge is performed again at PP7 for 3 minutes. PP8 changes the temperature inside the furnace from θ2℃ to θ3 with HEAT DOWN.
℃. After completing the 3 minute PP8 process,
Preparations for vapor phase growth are almost complete, and then proceed to PP9.
PP9 is EPi VENT1, 3 minutes, _H2 is
FH/min, dopant gas D _P FDPc.c./min,
Feed SiCl ₄ at a rate of 1 g FS/min. Gas _H2 is
PV2 → VC1 → PV3 → Chamber 106,
From the chamber 106, a gaseous mixture of SiCl ₄ and H ₂ flows from PV3 to VC1 and is supplied to the pipe PL6A. Chamber 104 to PV23, MFC
8. Dopant gas through MFC9 and MFC10
Since D _P is supplied, this gas D _P and the pipe PL
H ₂ + SiCl ₄ that has reached 6A merges and vents.
Ejected into VENT. Next, when it comes to PP10, this is FPi DEPO
The flow rate of each gas is PV19 as well as PP9.
is turned on. This continues for 3 minutes. Therefore, the aforementioned gas (D _P + H ₂ + SiCl ₄ ) in pipe PL1A is
PV19 joins and further enters the reaction path R1 through the main port of PV20, and the P is delivered to the wafer on the susceptor.
A type semiconductor is grown in a vapor phase. The growth reaction in this case is performed by a reversible hydrogen reduction reaction as described below. SiCl ₄ +2H ₂ Si+4HCl This is how Si accumulates on the wafer.
Phosphine PH ₃ is usually used as the dopant gas D _P. Note that B ₂ H ₆ (Diborane) is usually used as the dopant gas D _N for forming an N-type semiconductor. After 3 minutes have elapsed with PP10, the vapor phase growth is completed, and then with PP11, gas H ₂ is supplied for H ₂ purge at a flow rate of FH2/min for 3 minutes. Next, since PP12, PP13, and PP14 are not used, when PP15 is reached, HAET
OFF, cutting off the power supply to the dielectric heating coil. The reason why the time in this case is set to 3 minutes is to allow for the time required for the temperature inside the furnace to decrease. During this time, _H2 is also supplied at a flow rate of FH2/min. PP16 is a H ₂ Purge and lasts for 3 minutes. Next, move to PP17 and add N ₂ gas with N ₂ Purge.
FN17/min is maintained for 3 minutes. In FIG. 9, when a dopant is used to form an _N -type semiconductor layer,
From chamber 103 in Figure 7, PV5, PV7,
It is easy to understand that MFC4 and MFC5 are effective. In addition, in Fig. 7, when reactor R2 is used instead of R1, PV21 and PV22 are
It will be easily understood that it can be operated in place of PV19 and PV20. The above describes one embodiment of the apparatus of the present invention with respect to the specific structure of each of the plurality of reactors R1 and R2, the gas piping system, and the control contents of the process program. Next, parallel operation for efficiently operating a plurality of reactors, which is a feature of the apparatus of the present invention, will be explained. FIG. 10 is an explanatory diagram of the operation when the reactors R1 and R2 are operated in parallel. The parallel operation of the reactor in the present invention is performed in the reactor shown in FIG.
The process program can be executed in parallel in a part of the process program group for each reactor so that switching control can be performed so that the heating process of the reactor R2 starts at the same time as the heating process of the reactor R2 ends. This makes it possible to efficiently utilize heating power and shorten the cycle operation of each reactor. In FIG. 10, the upper side shows a graph showing the relationship between the furnace temperature θ°C of reactor R1 and the transition of each process program PP(i), and the lower side shows each graph related to reactor R2. process program
The transition of PP(i) is shown. In the same figure, reactor R
1, when the start switch S1 is turned on at time t0, the process program group PPG1 starts, and as shown in the figure, the process program group PP1,
When PP2 ends, it moves to PP3 and the induction coil is turned on.
The reactor R1 continues to be heated until PP15. The power supply to the induction coil on the reactor R2 side is as follows:
It is set to turn on immediately when PP15 on the R1 side ends. For this purpose, it is assumed that the start switch S2 for the furnace R2 is turned on when a time T4 has elapsed since the S1 switch was turned on. Now, let T3 (R2) be the time from when the S2 switch is turned on until the first process program PP1 of the process program group PPG2 starts, and further let T2 (R2) be the time from PP1 to the start of the same PP3. When driving, it is important to maintain the following relationship: T1 (R1) = T4 (R2) + T3 (R2) + T2 (R2). The S2 switch can be pressed at any time (generally, the S2 switch is pressed when the preparation work such as loading the wafer to be subjected to the vapor phase growth reaction into the furnace R2 is completed).
Turn on the switch. In the illustrated example, it is set as time t1. ) It is difficult to predetermine the actual start time t _S of PPG2. As a method for solving this problem, in order to actually perform parallel operation as shown in FIG. 10, it can be performed as shown in FIG. 11. In addition, in FIG. 10, the time interval T1
(R1) and T2 (R2) are known and pre-programmed time data. In FIG. 11, the register 131 has a time T
The difference between 1 and T2 is stored. On the other hand, the counter 132 receives pulses every second (second coefficient pulses).
Count from S1 switch ON to S2 switch ON. 133 is another counter,
When the S2 switch is turned on, the value of register 131 is T1.
A difference T3 between -T2 and the value T4 of the counter 132 is set, and after the S2 switch is turned on, this is subtracted by the second counting pulse, and when the count is out, PPG2 is actually started. Another method, although not shown, is to provide one counter and set T1, perform subtraction counting using a second counting pulse after the S1 switch is turned on, and then subtract T2 when the S2 switch turns on. If we further subtract it by the seconds counting pulse, we get
After all, when this counter reaches zero, PPG2
An actual start command may be given. FIG. 12 shows the data structure of a process program PP to be treated by the processing program of the present invention and a process program group PPG consisting of a series of the same process programs as A and B in the figure.
Figure A shows the contents of one process program, and the sequence number i of the process is stored in the first memory area. The next memory area shows the total data size of each memory area below. The data indicates the duration of this process sequence in minutes and seconds. Furthermore, an 8-bit data bit code is stored in the next memory area. A specific example is shown below. In the next memory area, the furnace temperature among the output data is set. Furthermore, the flow rate of gas H ₂ is set in the next memory area. Now, when temperature and gas _H2 are given as output data, the data bit code is 1 at the left end and the two bits to its right. Figure 12A is an example of one PP(i), but when there is no temperature specified as output data or when multiple types of various gases are output at the same time, each corresponding bit is set to 1, and the output data is The flow rate of gas to be sequentially output is stored in the memory area. Figure 12B shows the process program group PPG1, which corresponds to the parallel operation shown in Figure 10.
The data structure of PPG2 is shown. That is,
PPG1 is stored in the first memory area corresponding to reactor R1 on the left side. Hereinafter, a parallel operation time T1 and a parallel initial time T2 are set. In FIG. 10, this parallel initial time T2 is the sum of the sequence times of PP1 and PP2 in PPG1. PP1, PP after parallel initial time T2
2. Each process program data up to PP17 is stored, and then...
It becomes END OF PROGRAM. Figure 13 is a flowchart showing the processing process of the system program of this system (ROM2
4). In the figure, the first step after the program step start (hereinafter referred to as ST1)
Now, the setting value of the switch 67-4 in FIG. 5 or the input from the key input section 63, for example, S|
PROCESS.C, S｜RUN, S｜FETCH... The flags corresponding to the specified processing programs (some of which are indicated by symbols in the storage section 27 in FIG. 4) are stored in the RAM 27. Set in flag memory. Next, in ST2, the flag bit of the process control flag (S|PROCESS.C) is checked.
If YES in ST2, the process is controlled in ST3. If NO in ST2, move to ST4. Next, in ST4, reactors R1 and R2
It is checked whether each operation switch (for example, the reactor opening/closing pushbutton, motor rotation, etc.) is ON or not.
If it is ON, a command to execute the corresponding switch contents is given in ST5. If NO in ST4, proceed to ST6.
Each step from ST6 to ST25 is shown in Figure 4.
It corresponds to each processing program from MODiFY to TEST in ROM27. Normally, only one of ST6 to ST25 is selected and made valid. FIG. 14 is a flowchart showing details of process control corresponding to subroutine ST3 in FIG. 13. ST3 is attached to each processing step to indicate the specific content of ST3. In the same figure, if the start switch of reactor R1 is pressed in ST3-1, the process moves to ST3-2, where R1
It is checked whether the sequence remaining time of the currently operating process program PPi in the process program group PPG1 related to the process program PPG1 is 0. The remaining sequence time is determined by a certain number of seconds, as shown in the flowchart of the seconds processing program shown in FIG.
The sequence time of one PPi is loaded into a register and subtracted every second, and the result of this subtraction becomes the remaining sequence time. Now, in ST3-2, the remaining sequence time =
When it becomes 0, the next process gram PPi is specified in ST3-4, and its data is taken in. Next, in ST3-5, the sequence time of the newly designated PPi is set. As a result, the second processing program described above is activated. Next, in ST3-6, output data such as the flow rate of the specified gas and the temperature in the furnace are outputted, and further in ST3-7, data on various valves are outputted.
Give ON and OFF commands. Next, in ST3-8, it is incremented to i←i+1 in preparation for the next sequence set. In addition, the start speed of R1 is ST3-1.
When it is OFF, move to ST3-3, where reactor R
Check whether the start switch in step 2 is ON or OFF. If YES in ST3-3, further
Moving to ST3-9, it is checked whether the remaining sequence time of the currently valid process program of the process program group PPG2 related to R2 is 0. Below, the contents of each processing step from ST3-10 to ST3-14 correspond to ST3-4 to ST3-8. If NO in ST3-2, the process moves to ST3-3 via the confluence point. Also, ST3-3, ST3
If NO at -9, the process directly goes to END and ends the processing of this subroutine ST3. FIG. 15 is a flowchart showing details of subroutine ST7 in FIG. 13. For correction operations, please refer to S｜
When MCDiFY is keyed in, the system will issue a message saying MODiFY BELL-JAR=. Reactor R1, R2
Input 1→L, R2→R. Here L is
LEFT and R mean RiGHT. Next, as additional information, register the process name (PPGi, EPi, etc.) that will be formed after modification, creation date, creator, etc. Next, as a process pattern, input the type of process program and the type of process program group PPG to be modified. Next, the following message is issued from the system side, and the necessary information is entered and registered to modify one process program PP. The message corresponding to one PP and its modified state are shown below. SEQUENCE=PP3 N2=50 45 TiME=300 230 This example shows that the gas _N2 was modified from 50/min to 45/min and the time was modified from 3 min 00 sec to 2 min 30 sec. FIG. 15 shows details of the MODiFY processing flow. In the figure, the process program PPi to be modified is input in ST7-1. Proceeding to ST7-2, it is checked whether the input PPi belongs to the corresponding PPG.
If NO in ST7-2, it is checked in ST7-3 whether the modification operation is finished, and if YES, the subroutine ST7 ends. Also, ST7-3
If the answer is NO, an alarm command is given to the operator in ST7-4. If YES in ST7-2, the process moves to ST7-5, where time data (sequence time) is extracted from the output data of PPi and displayed. Next, the process moves to ST7-6, and if it is desired to process time data, the time is corrected as described above in ST7-7. As a result, the parallel operation time must be changed, so the calculation is performed in ST7-8, and the result is stored in the parallel operation time memory area of the PPG. Next, the process moves to ST7-9 and ST7-10, and if the temperature data is desired to be corrected, the correction is performed in ST7-11.
Furthermore, if it is desired to modify the gas flow rate in ST7-12 and ST7-13, the modification is performed in ST7-14.
Below are the necessary process programs in the PPG
It is now possible to specify PPi sequentially and make corrections. Note that to notify the system that the MODiFY operation has ended, enter S|. Next, subroutine ST9 in FIG. 13 will be explained. PROCESS is a processing function that generates a group of process programs to be executed. S｜PROCESS
When you enter , the system will send you a message. PROCESS BELL-JAR=Next, key in the reactor R1, R2 you want to use (L or R
Or enter L+R). Upon completion of input,
It would be fine if the system accepted it, but if you make an incorrect entry, an error message will be issued from the system.
The above-mentioned message is displayed again, and the user inputs the message again to ensure that it is entered correctly in an interactive manner. Next, as additional information, process name,
Register the creation date, creator, etc. as a directory. Next, input the type of process to be executed as a process pattern. If N type vapor phase growth is to be performed, the following display will be input. PROCESS PATTERN= EPi N (1) The underlined part is the input part. When various information is input in this way, a process program PP to be executed next is generated. Next, the system sends the following message, in response to which the necessary information is input, and the processes that configure EPiN are sequentially generated and completed. SEQUENCE= PP1 (2) N2= 50 (3) TiME= 300 (4) (2) is the sequence name, (3) and (4) are the necessary information attached to the sequence, and (3) If the gas used is _N2 gas, its flow rate is 50
The value is entered as (/minute). Continue
(4) is inputting the execution time required for the sequence. FIG. 16 shows S|RUN. S｜When RUN is entered, the currently running process (corresponding to one process program)
The latest information on each state change is displayed. If a process is running in reactor R1 or R2, the following will be displayed, for example: *RUN BELL-JAR=R1 SEQUENCE=EPi 1 DEPO SET TiME=1234 ××CURRENT DATA TEMP=1200 H2=55 DN=10 SiCl ₄ =22 TiME=59 (subtraction) (STOP TiME=×××× (increase)) ×RUN BELL−JAR=R2 SEQUENCE=N2 PURGE SET TiME=300 ××CURRENT DATA N2=55 TiME=59 Next, if one process has finished, S|S|
Display S|. The above-mentioned display contents are shown in the 16th section explained below.
The processing is performed by each processing step of the procedure shown in the flowchart of the figure. Each processing step in FIG. 16 will be explained below. First, in ST11-1, it is checked whether a process is being executed. ST11 if not running
Moving to -2, process end (S|S|S|) is displayed. If it is being executed in ST11-1, ST11-
3, the name of the process program PPi, which is the sequence being executed, is displayed. Next, ST11−
Moving to step 4, the sequence time (SET TiME) of the sequence PPi being executed is displayed. Next, ST11-5
Check whether the sequence is stopped at .
If YES, the process moves to ST11-7, and the cumulative stop time of the sequence PPi being executed is displayed. This sequence stop occurs when an operator issues a command to stop the sequence due to some abnormality during automatic operation, and normally sequence stop does not occur during normal automatic operation. Now,
If NO in ST11-5, the process moves to ST11-6 and displays the elapsed time TiME of the sequence being executed.Then, the process moves to ST11-8 and displays the execution sequence PPi.
Check whether the temperature is controlled.
If YES in ST11-8, the process moves to ST11-9 and displays the current measured temperature inside the furnace. If NO in ST11-8, the process moves to ST11-10 and checks whether the sequence PPi being executed is managing the gas flow rate. If YES in ST11−10,
In ST11-11, the current actual flow rate is displayed.If NO in ST11-10, the flow returns to the confluence point. Next, the processing program STEP shown in FIG. 4 will be explained. This processing program STEP is 1
It is prepared to change the order of each process program PPi in the process program group PPGi, and when you key in S | STEP, STEP BELL JAR= R1 is displayed, and the underlined part is reactor R1. is shown and the operator enters the key. Next, as additional information, necessary items such as the name of the process program group, creation date, creator, etc. are registered as a directory. Next, the system side asks which type of PPG sequential change is desired as a process pattern. This display is shown below. PROCESS PATERN= EPi N The underlined part is the operator's input part and indicates N-type vapor phase growth. Then from the system side, as shown below,
Each process program PPi in EPiN (one process program group) is sequentially shown as a SEQUENCE as shown below, so enter the necessary information using the keys on the right side of the equal sign. SEQUENCE=PP1 SEQUENCE=PP2 SEQUENCE=PP9 SEQUENCE=PP4 〓 〓 SEQUENCE=PP17 To end STEP, input S|. FIG. 17 shows details of subroutine ST13 (FIG. 13). In the figure, when S|STEP is input, a memory area for STEP is secured in ST13-1. Next, the process moves to ST13-2 and the sequence number i is input. Next, move on to ST13-3,
Here, it is checked whether the sequence number is i. If YES in ST13-3, the data of PPi is transferred to the reserved memory area.
If NO in ST13-3, move to ST13-5.
Check whether STEP is completed. If YES
STEP ends (END), if NO, ST13−
6, a warning is displayed to the operator that a non-existing sequence number exists, and this STEP process is ended. Next, STORE in the ROM 27 shown in FIG. 4 will be explained. The processing program STORE is used to transfer the generated (PROCESS) or modified (MODiFY) process program group PPG to the external storage medium of this system, such as a cassette magnetic tape, mini-floppy disk, or valve memory. . Figure 18 shows the subroutine ST15 in Figure 13.
A detailed flowchart is shown. In the same figure, ST15-1 is a process program group that is about to be stored in an external storage medium.
PPGk head process program (PPi, i=1)
Then, in ST15-2, the external storage medium (CMT51) used is initialized, and it is checked whether the device can be used. Next, the process moves to ST15-3, which is a determination step. In this step, it is checked whether all PPi's from RAM26 have been transferred to RAM53. If YES, RAM
This means that the transfer of data (for one PPGk) from 26 to RAM 53 has been completed, and then
Moving to ST15-4, data transfer from the RAM 53 to the CMT 51 is executed by the program in the ROM 52. When the transfer to CMT51 is completed,
In ST15-5, the operator is informed of the end.
On the other hand, if NO in ST15-3, ST15-
Move to step 6 and transfer the PPi data to RA53. Next, the process moves to ST15-7, and the presence or absence of an abnormality is checked, and if YES, the abnormality occurrence operator is notified in ST15-9. If NO in ST15-7, move to ST15-8 and start the next sequence PP
Increment i to i+1 to specify (i+1). When ST15-8 is finished, ST15-8 is completed again.
Return to step 3 and repeat this. As can be seen from the system block diagram shown in Figure 4, the STORE process runs the PPGk
, I/O bus 23, high-speed memory data transfer unit (HMT) 45, data highway 47, HMT5
0, the data is temporarily stored in the RAM 53 via the data bus 48. And stored in RAM53
PPGk is SUB-MASTER・CPU46・
The cassette magnetic tape 51, which is one of the external storage media, is stored by the ROM 52 via the cassette MTiF.
(CMT). Next, the processing program SORT shown in FIG. 4 will be explained. This SORT has the opposite effect to the above-mentioned STORE, and to explain it in Figure 4, it stores PPGk stored in the CMT51 in the RAM.
53 to the RAM 26. The subroutine in Figure 13 corresponding to SORT
A detailed flowchart of ST17 is shown in Figure 19.
In the figure, in ST17-1, the external storage medium in which a certain process program group PPGk to be SORT is stored is specified. Next, the process moves to ST17-2, and it is checked whether PPGk is actually stored in the external storage medium CMT51 specified in ST17-1. ST-2
If PPGk is not stored in
Move to ST17-3 and inform the operator that it is not stored. If YES in ST17-2, move to ST17-4, and move from CMT51 to RAM53.
(by ROM 52 and CPU 46). Then,
Proceeding to ST17-5, it is checked whether the transfer from the CMT 51 to the RAM 53 has been completed. If NO, return to ST17-4. If YES, move to ST17-6, where the data of each PPi is stored in the main memory.
Transfer to RAM26. Next, the presence or absence of an abnormality is checked in ST17-7, and if any abnormality is found, it is notified to the operator in ST17-8. If there is no abnormality, the process moves to ST17-9 and i is incremented to set PP(i+1) as the next sequence. Next, move on to ST17−10
Check whether the transfer from RAM53 to RAM26 is completed. If YES, it means that the SORT process is finished, and S|S|S| is displayed. Also
If NO in ST17-10, the process returns to ST17-6 and this loop is repeated. Next, the VERiFY processing program shown in FIG. 4 will be explained. This processing program once
The process program information stored in the RAM 26 is displayed on the CRT 29 for the operator to confirm. S｜When VERiFY is entered, VERiFY NAME= A message will be displayed, and by entering the process name in the underlined part, the corresponding process program group will be displayed.
The contents of PPGk are output. An example of the display output is as follows. PROCESS NAME=N N2 PURGF TiME=1234 FLOW N2=55 H2 PURGE 〓 N2 PURGE FLOW N2=55 In this way, the next service (display) is performed, the contents of N2 PURGE at the end of the sequence are output, and the process ends. Figure 20 shows the subroutine ST19 in Figure 13.
A detailed flowchart is shown. In the figure, in ST19-1, the first process program PPi of the current target PPGk is set. Next, the process moves to ST19-2 and the set PPi is set. Next, the process moves to ST19-2, and the sequence number i of the set PPi is checked.
If not, move to ST19-3 and check whether the VERiFY program is finished. If the process has not been completed, the process moves to ST19-4, where a warning is displayed to the operator regarding the appearance of a non-existing sequence number. If YES in ST19-2, the process moves to ST19-5 and takes in the time data (sequence time) in the data of PPi to the CRT RAM 25. Then,
Proceeding to ST19-6, the time data is displayed on the CRT29. Next, in ST19-7, check the presence or absence of temperature data in the PPi data, and if there is,
In ST19-8, the temperature data is displayed on the CRT29. Then, in ST19-9, PPi
The presence or absence of the flow rate data of the gas used is checked among the data, and if there is, the flow rate data is displayed on the CRT 29 in ST19-10. Next, move to ST19-11 and start the next sequence PP
Increment i for (i+1), return to the confluence point, and repeat ST19-2 to ST19-11 described above. Next, the processing program DiAGNOSiO shown in FIG. 4 will be explained. S｜When DiAGNOSiS is entered, the contents of the current process abnormality are provided (Fault Message). For example, as follows.

[Table] 〓 〓 〓

* FLOW ERROR N2 □□ □□
Here A0 indicates which flow controller,
A1 is the set flow rate, and A2 is the actual flow rate in the event of an abnormality. FIG. 21 1 shows the subroutine ST21 in FIG. 13.
This is a flowchart showing the details. In FIG. 1, ST21-1 checks whether there is any error registration. The controller registration is performed by checking the error code registration area (FIG. 3) of the RAMs 26 and 55 by executing the self-diagnosis program shown in FIG. 2. If there is an error registered in ST21-1, the process moves to ST21-2 and takes out the number of error registrations (K in Figure 3), then each error code is taken in ST21-3, and then the error code taken out in ST21-4. is converted into the corresponding fault message and displayed on the CRT. Next, the process moves to ST21-5 and the number of error registrations is decreased by one. Then, in ST21-6, it is checked whether the number of registered errors (the number of undisplayed errors) is zero, and if it is not zero, the process returns to ST21-3. Moreover, if it is zero, this processing program ends. 21. FIG. 21 shows the steps whose explanation was omitted in the explanation of the PROCESS/C processing program mentioned above.
This shows the details of the self-diagnosis function while the process is in progress, and the presence or absence of an abnormal state is checked in ST21-10. Items to be checked include temperature, flow rate, supply voltage to various valve devices, etc. If there is an abnormality in ST21-10, the process moves to ST21-11, and the necessary commands are given to release the abnormal state regarding the reactors R1, R2 or equipment in their surrounding areas. For example, this may be done by temporarily stopping the progression of the sequence itself. Next, in ST21-21, an alarm is displayed to give a warning to the operator. And ST21−
13, the error code is registered in the memory area corresponding to the RAM 26. Next, the processing program of USED TiME shown in FIG. 4 will be explained. USED TiME is a service that provides the usage time of each reactor R1 and R2 up to the present time, and the key input is S｜USED TiME.

[example]

S | CR NAME= CR OPEN (Y, *N) = CR or CR NO= CR NO= CR NO= CR EV5, 10 Open NAME= CR OPEN (Y, *N) = CR NO= CR NO= CR EV7 Open NAME=Call other functions MF function This is for operating the mass flow/paper riser controller for process control. NAME = CR INPUT (Y, *N) = Asks whether you want to perform an input operation or an output operation. For input operation, enter CR. NO = Mass flow number CR DATA = ×× ××: Data amount NO = Data at the time of input is serviced for only one minute with sequential number specification. Termination is NO=CR. For output operation, enter CR or CR. NO=Mass flow number CR DATA=Flow rate CR NO= Flow rate can be set by sequentially specifying the number. The end is CR. [example] S | TEST NAME= CR INPUT (Y, *N) = CR or CR NO= CR DATA= CR 15 specification (MFC5) NO= CR DATA= CR 300c.c. specification (MFC8) NO= CR NAME= CR INPUT (Y, *N) = CR NO = CR DATA = × × 1 minute service (MFC5) NO = CR DATA = × × 1 minute service (MFC8) NO = CR NAME = CRT function Add character to CRT display unit. When all the characters in the generator have been serviced and output for one page (80 characters x 25 lines) has been completed, the page is cleared and the process ends. NAME=CR Character service NAME=Call other functions LAMP function Turn on the lamps, LEDs, and BUZZER on the operation panels R1 and R2 and the control device panel in sequence.
When all output is finished, perform CLEAR, repeat this process 5 times, and then end this function. NAME= CR Output service NAME=Call other functions SOL function Switching SW, damper, R1 clamp, R1 lock, R1 seal, R1 exhaust, R2 clamp,
Turn on/off the solenoid valves used for R2 lock, R2 seal, and R2 exhaust. NAME = CR ON (Y, *N) = Ask whether you want to turn the valve ON or OFF. Input CR for ON operation. For OFF operation, input CR or CR. Next, NO=valve number CR executes the ON/OFF operation of the specified valve. ATC Function When the ATC function is called, the system automatically enters the local level. This function is used for temperature rise tests, soaking tests, etc. NAME= CR BELL-JAR= Which bell jar to use (R1, R2)
Ask a question. If a character other than
It is decided that the R2 side will be used in all cases. continue,
Gas system switched to _N2 purge line;
The Belgear begins its sealing action. Once sealed, the exhaust valve opens. At this time, the vacuum is
Not performed because it is an _N2 purge. In other words, the system is in a state where it is possible to execute processes on all selected bell gears. However, if an abnormal state is detected in the interlock, the ATC function will be canceled. 23 is a detailed flowchart for calculating the waiting time in the parallel operation shown in FIG. 10, and is one of the system programs stored in the ROM 24 of FIG. 4. In step ST1S in the figure, it is checked whether the start switch for reactor R1 is ON. If NO, move to ST7S.
If YES, the process moves to ST2S, where it is checked whether the other reactor R2 is already in process. If it is already being executed, the process moves to subroutine STS3 and calculates the waiting time of R1. On the other hand, if NO at ST2S, that is, the R2 start switch is not turned on yet,
The process moves to ST4S and checks again whether R1 is executing the process. If it is already running, the process moves to ST7S.
If NO in ST4S, the process moves to ST5S and sets a process control flag to execute the R1 process. Next, in ST63, PPi, that is, the first process program PPi of the process program group PPG to be executed in R1 from now on.
Set the sequence time to zero. Next, move on to ST7S. From steps ST7S to ST12S, by replacing R2 and R1, ST1S→ST7S,
ST2S→ST8S, ST3S→ST9S, ST4S→ST10S,
The processing contents correspond to each other, such as ST5S→ST11S and ST6S→ST12S. Up to now, we have explained Figs. 1 to 23, but in the above explanation, the reactor R1 or R2 has been explained as having a structure equipped with an induction heating coil, and we have explained how to effectively use the heating coil power supply during parallel operation. An example was explained therein. However,
Application of the control system according to the present invention is not limited to such an induction heating method, but can also be applied to, for example, a lamp heating method (FIG. 24). FIG. 24 shows a schematic configuration of a lamp heating type reactor. In the figure, reference numeral 201 is a supply section for various gases introduced into the reactor 207, which are introduced into the reactor 207 through a tube furnace 202 as shown by arrows. Reactor 2
A susceptor 209 is rotatably supported from above within the susceptor 209 and has a diameter increasing downward, and a semiconductor substrate 208 on which vapor phase growth is to be performed is placed on the outer peripheral surface of the susceptor 209. . A signal from the temperature sensor 210 is then given to a temperature controller 204 via a line 203. The controller 204 is the power converter 2
05 to provide a control signal to the power converter 205.
supplies power to the lamps 206 arranged around the outer wall of the furnace 207, so that the light emitted by the lamps 206 passes through the outer wall made of quartz or the like and reaches the semiconductor substrate 208. This lamp heating method heats the semiconductor substrate 208 directly from the lamp through the outer wall, and is also characterized in that it consumes less power than the induction coil method. By connecting multiple reactors using the lamp heating method to various gas sources, each reactor can perform processes independently without depending on the process status of the other reactors. It is possible to accomplish this. [Effects of the Invention] As is clear from the embodiments described above, according to the present invention, sequence time, gas type and flow rate, and temperature conditions are set as process parameters in a sequence process, and these are set as units of a process program. This method enables automatic movement of the equipment by appropriately combining the process programs of Compared to the control method, it is possible to significantly reduce the operator's manual work and decision making in setting the gas flow rate and temperature conditions, thereby providing an apparatus with excellent workability. Furthermore, in the apparatus of the present invention, the process program can be called up unit by unit according to the operator's request and displayed on the display means, making it possible to modify each process parameter and executing the contents in real time. It is possible to respond to the various operations required by the system as a sequence process adapted to actual physical and chemical conditions, and the contents can be grasped immediately. In particular, in the apparatus of the present invention, when a plurality of reactors are sequentially operated in cycles, the first reactor is By configuring so that a part of each process program of the process program group for the second reactor and the process program group for the second reactor can be executed in parallel,
The heating process is sequentially and continuously cycled for each reactor, making efficient use of heating power and shortening the cycle time, thereby achieving efficient operation of the semiconductor vapor phase growth apparatus. Therefore, according to the apparatus of the present invention, it is possible to programmably and quickly change the condition settings of process parameters in the preparatory stage of the optimal manufacturing process, and also to continuously and efficiently cycle multiple reactors. This not only improves the yield of products with stable quality, but also increases the operating rate of the equipment, and overcomes all the problems of conventional computer-based DDC control systems. A semiconductor vapor phase growth apparatus with extremely excellent control performance and productivity can be provided.

[Brief explanation of the drawing]

Fig. 1 is a diagram explaining the operation of a conventional vapor phase growth control device, Fig. 2 is a schematic configuration diagram of an embodiment of the device of the present invention, and Fig. 3 shows the flow of control information in the control system of Fig. 2. 4 is a block diagram of the vapor phase growth control system according to the present invention, FIG. 5 is a detailed front view of the panel of the control device shown in FIG. 2, and FIG. 6 is a main part of the reactor with a built-in induction coil. Cross section, 7th
The figure is a system diagram of the pipeline network connecting the reactor shown in Figure 6 and various gas supply sources, and Figure 8 is a system diagram of the pipe network connecting the reactor shown in Figure 6 and various gas supply sources. A system diagram showing the relationship between the mass flow meter and its surroundings, FIG. 9 is a format diagram of an operation table showing flow rates in the pipe network, and FIG.
Time chart when running R2 in parallel,
Figure 11 is a block diagram of the time calculation section during parallel operation, Figures 12A and 12B are format diagrams explaining the data structure of a process program and a group of process programs, and Figure 13 shows the processing process of the system program of this system. 14 is a detailed flowchart of subroutine ST3 in FIG. 13, and FIG. 15 is a subroutine in FIG. 13.
A detailed flowchart of ST7, FIG. 16 is a detailed flowchart of subroutine ST10 in FIG. 13, FIG. 17 is a detailed flowchart of subroutine ST13 in FIG. 13, and FIG. 18 is a detailed flowchart of subroutine ST15 in FIG. 13. , Figure 19 is the first
Detailed flowchart of subroutine ST17 in Figure 3,
20 is a detailed flowchart of subroutine ST19 in FIG. 13, FIG. 21 is a detailed flowchart of subroutine ST21 in FIG. 13, and FIG. 22 is a flowchart of a seconds processing program related to subroutine ST23 in FIG. 13. FIG. 23 is a flowchart for calculating waiting time, and FIG. 24 is a diagram showing the main part of a lamp heating type reactor. 11... High frequency generation section, 12, 13... Reactor, 14... Control section, 16... Equipment drive section, 21
...CPU, 22...Data bus, 23...i/
O bus, 24...Storage section, 25...CRT RAM,
26...Temporary storage unit, 27...ROM, 28,4
9...Interface, 29...CRT, 30,
32...Input module, 31...Keyboard,
34, 36... Output module, 35... Output section, 37... Gas valves, 39... Relay drive section,
40...Motor and valve, 41...D/A conversion module, 43...A/D conversion module, 46
...CPU, 47...Data highway, 48...
...data bus, 45, 50...high-speed memory data transfer unit, 51...magnetic tape, 52...ROM,
53...RAM, 54...i/O bus, 61...
Display device, 62...Cassette tape mounting section, 63...Key input device, 64...Temperature control section, 67...Data input area, 71...Bottom plate,
72...Gas inlet, 73...Exhaust hole, 74...
Pipe line, 75... Susceptor, 76... Rotating member, 7
7... Motor, 78... Cover, 79... Induction heating coil, 80... Insulating plate, 81... Bolt,
82, 83... Connection joint portion, 84... Quartz layer, 8
5...First stainless steel layer, 86...Second stainless steel layer, 87...Brim, 88...Clamp member, 9
0...Ceiling cover, 91...Wafer, 92...Observation window, 93...Temperature detection window, 101, 102, 10
3,104,105...Gustyamba, 106...
... Bubbling chamber, 121 ... A/D converter, 122 ... Analog multiplexer, 123
...D/A converter, 131 ...Register, 13
2,133...Counter.

Claims (1)

[Scope of Claims] 1. A plurality of reactors for performing vapor phase growth on substrates such as silicon, means for heating each of the substrates, various gas sources necessary for vapor phase growth, and each of the reactors. a network of pipes connected to each other; a valve device provided on the network to form the network so as to guide desired amounts of the various gases to each reactor; and an on/off switch for the valve device. In a semiconductor vapor phase growth apparatus, the heating means is arranged for each substrate placed in each reactor, in a semiconductor vapor phase growth apparatus comprising a signal for controlling off or opening thereof and a control device for giving a signal for controlling each of the heating means. At the same time, a changeover switch for selectively connecting each of the heating means to a common power source for supplying heating power, and a start switch for starting each of the reaction furnaces, respectively; (b) the time from when the first start switch of the first reactor is energized to the end of heating of the corresponding first process program group; c) the time from the start of the process to the heating start command of the second process program group corresponding to the second reactor following the first reactor, and (d) the time from the start of the process to the heating start command corresponding to the second reactor From the time when the start switch No. 2 is activated, the heating power supply to the second reaction furnace is started after the heating power supply to the first reaction furnace ends. A semiconductor vapor phase growth apparatus characterized by comprising a calculation section for setting a process sequence start time. 2. In the semiconductor vapor phase growth apparatus according to claim 1, the calculation unit (e) calculates the execution time from the start of each process to the end of heating of the first process program group corresponding to the first reactor. (f) at the same time as the process starts, the contents of the counter are decremented at regular time intervals; and (g) the value of the counter is decremented from the value of the counter in response to activation of the second start switch. subtract the time from the start of the process to the heating start command in the second process program group corresponding to the second reactor, and (h) when the remaining time is further subtracted, the second process program A semiconductor vapor phase growth apparatus configured to command the start of execution of a group. 3. In the semiconductor vapor phase growth apparatus according to claim 1, the calculation section (i) when starting execution of the first process program group corresponding to the first reactor, performs the following steps from the start to the end of heating. A counter is provided to set a value T ₁ −T ₂ obtained by subtracting the time T ₂ from the start of the process of the second process program group corresponding to the next second reactor to the heating start command from the time required T ₁ to reach the heating start command, (j) Decrease the value of the counter at regular intervals after the first process program group starts executing, and (k) instruct the start of execution of the second process program group when the counter value becomes zero. A semiconductor vapor phase growth apparatus configured as follows.