Search

Ion implantation requires a high beam current stability to ensure a uniform implantation dose across the wafers. This requires ion sources with a stable extraction system to achieve a high beam current stability. Our goal is to extract 0.5 mA of C2- with less than 10 glitches per hour, with a glitch defined as any variation greater than +/- 5% of the nominal beam current. The beam energy should be between 10 keV and 30 keV while the normalized 4 RMS emittance should be less than 1 mm·mrad. We’ve shown that high negative carbon ion current densities could be obtained with a multi-cusp ion source when acetylene was used as the feed gas (0.27 mA of C2-), but there was significant sparking between the electrodes, which led to frequent beam current glitches (>1000/hour). In this study, we investigate the different factors that contribute to the sparking between the electrodes. We found that the sparking is highly correlated to the dumping of the co-extracted electrons and the beam strike on the electrodes. Extraction simulations were completed to determine how the electron dumping can be improved and how the beam strikes on the electrodes can be reduced. Furthermore, we modified the magnetic configuration in the plasma chamber to decrease the electron density close to the extraction aperture, which reduced the co-extracted electron current by a factor of almost 5 and reduced sparking frequency to about 20 glitches per hour. However, there was a corresponding decrease in the extracted beam current due to the smaller aperture sizes needed to reduce the sparking, with only 0.05 mA of C2-.

D-Pace’s 13.56 MHz RF powered H− ion source, a hybrid design between the TRIUMF licensed filament powered volume-cusp ion source and the University of Jyvaskyla licensed RF ion source, has been shown to be less efficient than the filament powered ion source, even though both sources use the same body and extraction system. The difference is thought to be due to RF power losses to the outside of the ion source, to the lack of plasma confinement on the back plate of the ion source, and to the absence of a sputtered tantalum coating on the plasma chamber walls. We believe that the lack of confinement on the back plate also causes the RF window to heat, with a maximum temperature measured at 450 °C at 3.5 kW of RF power. In this paper, we are investigating the use of a solenoid and permanent magnets behind the antenna and the back plate of the ion source to create a magnetic field that confines the plasma by preventing the electrons from striking the RF window. Furthermore, we present the effect of sputtering a tantalum coating in the plasma chamber of the ion source on the production of H− ions in the RF powered source. Our results show an increase in H− beam current at higher RF powers with a fresh coat of tantalum, and a subsequent decrease in beam current over time.

Acetylene and carbon dioxide gases are used in a filament-powered volume-cusp ion source to produce negative carbon ions for the purpose of carbon implantation for gettering applications. The beam was extracted to an energy of 25 keV and the composition was analyzed with a spectrometer system consisting of a 90° dipole magnet and a pair of slits. It is found that acetylene produces mostly C−2 ions (up to 92 µA), while carbon dioxide produces mostly O− with only trace amounts of C−. Maximum C−2 current was achieved with 400 W of arc power and, the beam current and composition were found to be highly dependent on the pressure in the source. The beam properties as a function of source settings are analyzed, and plasma properties are measured with a Langmuir probe. Finally, we describe testing of a new RF H− ion source, found to produce more than 6 mA of CW H− beam. peerReviewed