Energy Linearity Calibration Method for High Voltage Power Supply of Medium to High Energy Ion Implanter
Ion implantation at medium to high energies requires precise control of the ion energy to achieve accurate implant depth profiles. The acceleration voltage determines the ion energy, and any nonlinearity between the commanded and actual voltage causes errors in the implant depth. Energy linearity calibration characterizes and corrects for this nonlinearity, enabling accurate implantation across the full energy range of the implanter.
Medium to high energy ion implanters operate at energies from hundreds of kiloelectronvolts to several megaelectronvolts. These energies are achieved through multiple acceleration stages, including extraction from the ion source, post acceleration, and possibly a high energy acceleration column. The total ion energy is the sum of the voltages applied to each stage. Precise control of each voltage is essential for accurate energy determination.
Energy linearity refers to how well the actual ion energy corresponds to the commanded energy. In a perfectly linear system, the actual energy is exactly proportional to the commanded value across the entire operating range. Nonlinearity causes the actual energy to deviate from the commanded value, with the deviation varying across the energy range. The nonlinearity must be characterized and corrected for accurate implantation.
Sources of nonlinearity include several mechanisms in the high voltage power supplies and the acceleration column. The digital to analog converter that sets the voltage command may have nonlinearity in its transfer function. The feedback divider resistors may have voltage coefficients that cause the division ratio to vary with voltage. Space charge effects in the acceleration column can cause voltage depression that varies with beam current. Magnetic field effects from solenoids or other elements can affect the effective acceleration.
Calibration methods for energy linearity involve measuring the actual ion energy at various commanded energies and determining the correction needed to achieve accurate energy. Several techniques are used for energy measurement in ion implanters, each with advantages and limitations.
Magnetic analysis uses a magnetic field to bend the ion beam, with the bend radius depending on the ion momentum and thus the energy. A Faraday cup or other detector measures the beam position after the magnet. The relationship between beam position and energy enables energy measurement. Magnetic analysis is highly accurate but requires a dedicated magnet and detector system.
Electrostatic analysis uses an electrostatic field to deflect the ion beam. The deflection angle depends on the ion energy and the applied field. This technique can be implemented with parallel plate or cylindrical analyzers. Electrostatic analysis is simpler than magnetic analysis but may have lower resolution.
Time of flight measurement determines the ion energy from the time required to travel a known distance. This technique is most applicable for pulsed beams where the time of emission is known. The flight time is inversely related to the velocity, which is related to the energy. Time of flight can be very accurate for appropriate beam conditions.
Nuclear reaction analysis uses the energy dependence of nuclear reaction cross sections to determine ion energy. A target with a known resonance reaction is placed in the beam path. The reaction yield varies sharply with energy near the resonance. By measuring the yield as a function of commanded energy, the actual energy can be determined. This technique is highly accurate but requires appropriate nuclear reactions.
Implant depth measurement provides a practical calibration method for production implanters. Test implants are performed at various commanded energies, and the resulting depth profiles are measured by secondary ion mass spectrometry or other techniques. The measured depth profiles are compared to the expected profiles based on the commanded energies. The discrepancy reveals the energy error, which can be used to determine the nonlinearity correction.
The calibration data are used to generate a correction function that maps commanded energies to corrected commands that produce the desired actual energies. The correction function can be implemented as a lookup table with interpolation, or as a polynomial or other mathematical function fitted to the calibration data. The correction is applied in the implanter control system when setting the acceleration voltages.
Regular recalibration maintains the accuracy of the energy linearity correction as the system characteristics change over time. Component aging, contamination, and maintenance activities can affect the energy linearity. The recalibration frequency depends on the stability of the system and the accuracy requirements of the application. Statistical process control monitors implant results to detect when recalibration is needed.
The uncertainty of the energy calibration contributes to the overall uncertainty in the implant depth. Error propagation analysis determines how calibration uncertainties translate to depth uncertainties. The calibration method and correction function should be designed to minimize the contribution to implant uncertainty within practical constraints of calibration time and complexity.
