Integrated Dynamic Simulation for Process Optimization and Control

Transcription

1 Interated Dynamic Simulation for Process Optimization and Control G. Brian Lu, Laura L. Tedder, Monalisa Bora, and Gary W. Rubloff SF Enineerin Research Center for Advanced Electronic Materials Process orth Carolina State University, Raleih, C Outline: Motivation Process/Equipment Analysis, Desin, and Control Simulator Structure Dynamic and Physical Approach PolySi RTCVD Simulator Structure Dynamic Characteristics Simulator Validation Sensor Sinals, Reaction Kinetics Film Deposition Applications in Process Optimization for Manufacturin and the Environment Process Cycle Time Materials Utilization Efficiency Other Manufacturin Fiures of Merit Applications in Manufacturin Education and Trainin Conclusions Acknowledments: ational Science Foundation Semiconductor Research Corporation Visual Solutions, Inc. (VisSim consultations)

2 Motivation: Both sinle-wafer and batch semiconductor processes are discrete, fairly short steps Transient / dynamic characteristics of process influence properties and reproducibility of product Finite times needed to establish or chane process conditions (e.., start and finish) Variations durin nominally steady-state process conditions Transitions between different chemical reaction reimes Dynamics may be a crucial factor in desinin efficient processes, equipment, sensor, and control systems

3 Dynamic Simulators for Sensor-Based Process Control Process Recipe action vs. time Process Equipment Equipment Simulator Process Simulator real data Sensor Sensor Simulator virtual data Comparison for Control Response Advisory Information Run-to-Run Control Real-Time Control ldynamic simulation to capture time-dependent behavior lsimulators ==> virtual sensor data lcomparison of real and virtual sensor data ==> control

4 Dynamic Simulation Applications Process/equipment desin and performance analysis Sensor sinal simulation Fault simulation and classification ew process recipe ew equipment confiuration Process extension to new parameter reimes Flexible manufacturin Manufacturin optimization Process cycle time Environmental optimization Materials utilization efficiency

5 Our Approach to Dynamic Simulation Use physically-based models wherever possible Radiative heat transfer Mass balance Boundary layer transport Surface adsorption/reaction/desorption kinetics Growth rate, film thickness Use other models elsewhere to complete the system-level description Reduced-order models to represent hih complexity (e.., reactor fluid dynamics, heat transfer) Empirical/statistical models where physics/chemistry poorly known (e.., plasma process) Time-dependent behavior of the equipment and process Real-time based functions for all input and output parameters Interation of simulators for equipment, materials/process, sensor, and control system Linked equipment/process/sensor/control simulators throuh lobal parameter variables Control system implementation (sequential event driver)

6 Schematics of Polysilicon RT-CVD Reactor heatin lamps SiH 4 Si + 2H 2 MFC 1st stae 2nd stae QMS as cylinder RTP reactor pumps pumps meterin valve RTP pumps 1.0 mm samplin aperture Gas Handlin RTP Reactor and Pumps Mass Spec Samplin

7 Windows-based RTCVD Simulator Structure Gas Handlin RTP Reactor and Pumps Mass Spec Samplin Compound Block Structure Second Level Compound Block

8 Schematics of a VisSim Block-diaram and the Correspondin Mathematical Expressions F dt A A B A out + = A A B + F A A B + F dt A A B + A B +

9 Dynamic RTCVD PolySi Simulator Process Recipe Valves, MFC s vs. time, status Lamp power vs. time Overall process timin, conditions Equipment Simulator Gas Flow Vacuum chambers Mass flow controllers Pumps, valves Conductances, volumes Partial and total presssures Pressure control system Viscous/fluid flow partial press s Heat Flow Wafer absorptivity, emissivity Wafer thermal mass Wafer radiation, conduction Wafer temperature Temperature control system Process-dependent absorptivity, emissivity Convective heat loss in fluid flow temperatures Process Simulator Sensors Sensors and and Control Control System System Sensors Total and partial pressures Temperatures Valve and MFC status Controls PID controlers for temperature and pressure Lamp power output control Throttle valve positions CVD Reaction Gas phase transport Boundary layer transport Surface-condition-dependent reaction rates - surface kinetics rates Wafer State Deposition rate Film thickness Thickness control system Product properties - uniformity, conformality, material quality, toporaphy, reliability Manufacturin FoM Simulator Manufacturin Process Efficiency Cycle time Consumables volume Enery consumption Environmental Assessment Gaseous emissions Reactant utilization Power consumption Solid waste

10 Examples of Dynamic Elements Time and history dependent film deposition: Temperature and partial pressures vary continuously throuh process Surface reaction and deposition rates vary continuously throuh process Film thickness depends on detailed process history Process-dependent absorptivity/emissivity and temperature control: Film thickness chanes durin deposition Absorptivity/emissivity chanes with film thickness Temperature variation is compensated by PID control of lamp power output PID pressure control PID throttle valve controller stablizes initial reactor pressure Surface reaction induced pressure chane is compensated by PID control of throttle valve position.

11 Go to landscape file for this fiure Film Thickness (A) Ar SiH 4 H 2 QMS Partial Pressures wafer T ( o C) rowth rate (Α/s x10) Temperature and Rate

12 Direct Output from the Simulator Process time-profile: Partial and total pressures in reactor Wafer temperature and optical properties Reaction rates Deposition rate Equipment time-profile: Valves and MFCs Lamp power Control system Sensor sinals Manufacturin Fiures of Merit Film thickness (and quality) Process cycle time and throuhput Materials consumption Power consumption Environmental Fiures of Merit Gas by-product formation Partial pressures in down-stream as Materials utilization efficiency Process mass balance

13 Comparison of PolySi RTCVD Film Thickness Between Experimental and Simulation 9000 Thickness from Simulation (A) C 640C 650C 675C 700C 725C 750C 775C 800C 850C Thickness from Experiments (Α) Simulator captures dynamic physics and chemistry well to first order Some systematic subtleties not yet represented

14 Process Optimization for Environmentally Conscious Manufacturin To identify where sinificant ains are possible To minimize experiments needed for optimal process desiin Manufacturin Fiures of Merit Film thickness (and quality) Process cycle time and throuhput Materials consumption Power consumption Environmental Fiures of Merit Materials utilization efficiency Power consumption Partial pressures in down-stream as Process mass balance

15 Dynamic Simulation for Environmental and Manufacturin Optimization RTCVD PolySi: timin and temperature dependence, 300 sccm 50 SiH 4 Utilization (%) o C 700 o C 650 o C Process Cycle Time (sec) o C 700 o C 750 o C Pressure (torr) at which heatin beins start as flow and heatin simultaneously start heatin after as flow established Hiher temperature and earlier wafer heatin benefit BOTH manufacturin and environment: hiher SiH 4 utilization shorter process cycle time Dynamic simulation permits quantitative prediction for complex behavior

16 Dynamic Simulation for Environmental and Manufacturin Optimization RTCVD PolySi: timin and flow dependence SiH 4 Utilization (%) sccm 200 sccm 300 sccm 450 sccm 600 sccm 1000 sccm Process Cycle Time (sec) sccm 200 sccm 300 sccm 450 sccm 600 sccm 1000 sccm Pressure (torr) at which heatin beins start as flow and heatin simultaneously start heatin after as flow established Lower flow rate and earlier wafer heatin: hiher SiH 4 utilization (reduced consumables cost, environmental benefit) somewhat hiher process cycle time Dynamic simulation permits quantitative prediction for complex behavior

17 Manufacturin Optimization manufacturin fiures-of-merit (FoM s) materials reliability material quality equipment CoO consumables cost environment cycle time equipment desin ramp rates control parameters as flow timin pressure temperature equipment and and process parameters and and desin desin Time-dependent physical and chemical behavior determines manufacturin fiures-of-merit (FoM s) Multiple manufacturin FoM s result from equipment, process, control choices and parameters

18 Dynamic Simulation Applications Equipment desin and performance 3-stae CIS mass spec vacuum system dynamics Process recipe desin Interpolation and extrapolation for multi-parameter process models Sensor sinal simulation Linked equipment/process/sensor simulators Virtual sensor sinals to compare with actual sinals Manufacturin optimization Process cycle time, consumables cost Environmental optimization Reactant utilization efficiency Manufacturin education and trainin Modules providin real-time, interactive operator trainin Desin for contamination control Gas impurity variation with use from liquid source Control systems behavior PID controls for temperature and pressure Fault simulation Equipment/process/sensor failure and consequences Dynamic Simulator based equipmemt/process control

19 Simulator Module and Interation for Manufacturin Trainin and Education Dynamic simulator interated with HELP files to form stand-alon simulator module Simulator modules incorporated into manufacturin education and trainin courses Simulator Module - MOCVD System Module HELP file Dynamic Simulator Dynamic Simulator HELP file Dynamic Simulator: MOCVD SYSTEM Course lectures Dynamic Simulation Exercise

20 Conclusions Crucial issues for semiconductor manufacturin demand understandin and analysis of complex, multi-variable behavior as a function of time Commercial software provides a powerful, versatile vehicle for dynamic simulation Dynamic behavior of equipment, process, sensor and control systems are replicated in real-time with physically-based models umerous important applications to manufacturin exist Equipment desin and performance Process recipe desin Sensor sinal simulation Control systems behavior Manufacturin optimization Environmental optimization Manufacturin education and trainin Fault simulation Desin for contamination control