Rapid thermal annealing (4-7s) of 28Si and 9Be implants in VPE-grown In 0.53Ga 0.47As has produced n-and p-type active layers with controlled doping levels between 1017 and 3x108 cm-3.The multiple-implant schedules were based on R and AR data derived from SIMS measurements on p P single-energy implants. The activated n-and p-type layers have a good surface morphology and 300 K mobilities of 3000-7000 and 100-200 cm 2 /V-s, respectively. Data on implant schedules, electrical characteristics, carrier concentration profiles, and Rp /AR information are presented.

Wafer charging in high‐current ion implanters is reviewed, and the operation of the electron flood gun in the Varian 160‐10 implanter is examined. It is shown that flood gun electrons with energies up to 350 eV do reach the wafers and can cause damage when wafers are excessively overflooded. An in situ flood gun monitor using a capacitive pickup sensor is described. Experiments with the capacitive charge sensor have further shown that (i) wafers can self‐charge during pumpdown or venting of the target chamber, (ii) a slight overflooding is preferable to under flooding, and (iii) for perfect neutralization the flood gun emission current should vary with the magnetic scanning of the ion beam across the wafers. Using metal oxide semiconductor capacitors as test vehicles, it is shown that other factors also can affect charging damage to devices during implantation, such as the thickness of the field oxide or photoresist relative to the ion penetration depth, the proper grounding of the back sides of wafers during implantation, and the polarity of the silicon underneath the gate oxide. Finally, the benefits of proper electron flood control are demonstrated, and operating procedures are recommended.

No abstract

No abstract

The problem of low‐energy B+ contamination in the implantation of B++ ions using pre‐analysis type ion implanters is examined and the causes of contamination mechanisms are discussed. Experiments were performed on pre‐analysis‐type medium‐current implanters to investigate the effects on B++ beam purity due to the use of solid boron source and liquid boron source in an effort to lower ion source chamber operating pressure, the use of electrostatic ion beam filter to remove low‐energy contaminants from the dissociation of B2+ molecules if any, the use of aluminum waveguide liner to reduce gas desorption, and the effects of improved vacuum after cryopump regeneration. Experimental results indicated that, despite all these improvements, uncontaminated B++ implants cannot be assured for either BF3 gas or solid LiBF4 source. The secondary electrostatic analysis is demonstrated to be a sensitive technique for setup and monitoring of beam purity of doubly charged ion implants, and essentially uncontaminated B++ implants have been achieved with either BF3 gas source or the LiBF4 solid source. This was previously impossible without the refined secondary electrostatic analysis technique.