Handbook of Modern Coating Technologies

Experimental results obtained with the use of pulsed beams

Fig. 5—16 presents typical average positron lifetimes depending on implantation energy E and penetration depth z1(E). The different shapes of the curves in Fig. 5—16A are due to back diffu-sion and surface effects. Two n-type Si specimens with an impurity concentration of 1015/cm3 were investigated. The mean positron lifetime тт for both samples at implantation energies higher than 5 keV was practically identical and assumed to be constant. The values of тт dif-fered significantly at lower energies. The measured lifetime spectra for Si specimens from which the surface oxide layer was stripped off were amenable to analysis only with the use of two lifetime components, the intensities of which are 385 and 225 ps, respectively. The larger component intensity decreased roughly to 75% (to 10% at an implantation energy of ~18 keV) because of the probability of positron back diffusion toward the surface. A lifetime of 385 ps is, as a rule, indicative of positron annihilation near the surface [140,160—164]. Single-crystalline silicon without defects is characterized by a shorter lifetime, 225 ps. The mean lifetime in the oxide layer at the surface of an Si single crystal under low implantation energies increases to 1650 ps. The analysis of measurements of the lifetime revealed the following components: 1350, 450, and 225 ps with appropriate intensities. Evidently the largest lifetime is characteristic of the formation of positronium (Ps) in the oxide layer. The value of 450 ps corresponds to the presence of pores in oxide layer and at the interface between oxide and single crystal. The 225 ps positron lifetime is attributable to the bulk silicon single crystal. Fig. 5—16D and E depicts positron lifetimes for 4H-SiC specimens differing in the degree of doping and the thick¬ness of the oxide layer (30 and 70 nm, respectively). 
0 4 8 12 16 20
Positron implantation energy (keV)
FIGURE 5-16 (A) Average lifetime of the positron in Si of n-type with and without oxide layer plotted versus positron implantation energy and mean penetration depth. (B) Average lifetime of the positron in SiC of p- and n-type with different thickness oxide layers plotted versus positron implantation energy and mean penetration depth. (C) Dependence of the average lifetime of the positron in SiC of p-type after implantation of Al and subsequent annealing on the mean penetration depth. For comparison the results of as-grown specimens are given. (D) Dependence of the shortest lifetime (T1) in Al+-doped SiC samples on positron implantation energy. Lifetime of the second component T2 5 480 ps. (e) Dependence of average positron lifetime in p-type SiC shown in (B) on positron implantation energy and mean penetration depth [47].
In doped n-type specimens, the diffusion toward the surface is similar to that in doped p-type SiC samples, which can be explained by the negative charge of free carriers. It is espe-cially apparent in an energy range of 4—12 keV. In the oxide layer in both samples formation of positronium is observed. For the latter sample, the spectrum of lifetime was adjusted (approximated) to fit the three lifetime components with the respective intensities. The results are given in Fig. 5—16C.
An orthopositronium lifetime signal from the detector was observed only in the 70-nm thick oxide layer; its intensity rapidly decreased with decreasing implantation energy. The life¬time component T2 corresponds to vacancy type clusters (most likely microtubes) produced in the course of crystal growth. The intensity I2 of the second lifetime component remains low (roughly 10%), notwithstanding high implantation energies. The shortest lifetime T2 is charac-teristic of bulk single crystals and decreases due to positron trapping in defects and surface states. One of the promising materials for nuclear radiation detectors operating at high 
temperatures is SiC [163]. However, a local modification of the doping intensity is crucial for the production of such devices. The long range of impurity diffusion necessitates the use of high treatment temperatures exciding 2300K that cause decomposition of SiC [164,165]. Ion implantation may be an alternative modality, since it allows low-temperature processing and plate structuring [166]. However, these procedures do not prevent the formation of radiation defects. The postimplantation defects were studied after their annealing by measuring lifetimes for various 4H-SiC specimens in the form of a 4-mm thick epitaxial layer of p SiC grown on an SiC n-type substrate with a degree of doping equal to 7.3 X 1018/cm3. Al impurities were implanted at five different energies and a total fluence of 2.35 X 1013/cm2 to obtain a uniform implantation profile. The results are demonstrated in Fig. 5—16E.
The average positron lifetime after implantation and annealing was compared with the results of measurements in a SiC bulk crystal with p-type conductivity. The average lifetime тт in this sample increased at the lowest implantation energies due to annihilation in the surface states. As the energy increased, тт rapidly dropped to 141 ps;such behavior is typical of bulk single crystals and agrees with theoretical predictions.
The results of measurements for Al1-doped samples were altogether different. The life¬time was 218 ps at energies of positron from 2 to ~10 keV (Fig. 5—16E), suggesting satura¬tion trapping in the defects resulting from the implantation of Al1 ions.
The results of these measurements are very similar to the calculated lifetime value of ~216 ps obtained for divacancies VSi, Vc [67]. The lifetime т1 shown in Fig. 5—16E is unusual for the bulk single crystals. In all probability, it is the integrated result of the measurements (signals) obtained for the bulk material and induced defects. The defective and implanted regions fully recovered after annealing. The lifetime of annihilated positrons (494 ps) is slightly longer than that in as-grown samples (480 ps). The cause of тт growth at an energy above 10 keV remains to be elucidated. This feature was observed only in epitaxial samples and may be attributed to the peculiarities of growth and/or p—n junction effects.