Implant annealing temperatures

Figure 4.15 Plot of calculated silane flow rate versus implant annealing temperature for various temperatures in the range of 1,400-1,850°C. Dotted line indicates the empirical reference value of 6 seem at 1,600°C.

As discussed in the introduction, implant annealing temperatures well above 1100°C will be needed to fully remove the implantation-induced damage in GaN. Initial studies of redistribution of impurities at... 

Ion implantation has been successfully used to dope the IITSb material system. Sulfur has been used as an n-ty e dopant, although with poor activation efficiencies (175). -Type doping has been achieved using beryUium, zinc, and magnesium (175,176). Activation of the -type dopants is generally much better, near 50%. For the Sb-containing materials the post-implant anneal is conducted at much lower temperatures, typically <600° C. 

FIG. 77. Room-temperature photoluminescence spectra of Er-implanted PECVDa-Si H, annealed at 400°C. The implantation energy and dose were 125 keV and 4 x 10 Er/cm". respectively, which resulted in peak concentration of 0.2 at.%. "Low-0" and "high-O" denote a peak oxygen concentration in a-Si H of 0.3 and 1.3 at.9c. respectively. The inset shows the 1.54-/im peak intensity as a function of annealing temperature, for both oxygen concentrations. (From J. H. Shin. R. Serna, G. N. van den Hoven, A. Polman, W. G. J. H. M. van Sark, and A. M. Vredenberg. Appl. Phys. Lett. 68. 697 (1996). 1996, American Institute of Physics, with permission.]... 

Annealing temperature and symmetry of several H+ implantation induced... 

A silane-based CVD reactor suitable for performing high-temperatnre anneals in an Si- rich ambient was used for these experiments [86]. The samples were placed on a SiC-coated graphite susceptor and an RF induction coil used to heat the susceptor to temperatures on the order of 1,600-1,800°C. Silane and argon were the two process gases used, where Ar not only serves as a dilutant gas but also as a carrier gas to transport silane molecules to the crystal surface. All the implant annealing experiments were performed at atmospheric pressure. 

The final implant annealing process schedule developed during this research is shown in Figure 4.19. A 6-slm UHP Ar flow is first established in the reactor. When the RF generator is turned on, the susceptor is heated to the annealing temperature (typically 1,600°C) using a controlled thermal ramp. To avoid the formation of Si droplets, silane is not introduced into the reactor until a substrate temperature of 1,490°C is reached. At that time the premixed silane in Ar gas is introduced into the Ar carrier flow at a flow rate of 20 seem. All flows are controlled using calibrated... 

Figure 4.19 Implant annealing process schedule indicating gas flow timing versus sample temperature.

Making comparisons between literature data using different measurement techniques is therefore many times not possible. In addition, there is the problem with surface decomposition at higher annealing temperatures, discussed in Section 4.3.1, that may strongly affect the formation and reproducibility of electrical contacts produced on implanted and annealed material. In this section we will nevertheless try to evaluate and compare recent achievements in this important field and describe a selected number of recent results on activation studies on both donors and acceptors. 

For implanted acceptor activation there have been several reviews during the last few years since Troffer et al. s often-cited paper on boron and aluminum from 1997 [88]. Aluminum is now the most-favored choice of acceptor ion despite the larger mass, which results in substantially more damage compared with implanted boron. Mainly it is the high ionization energy for boron that results in this choice, as well as its low solubility. In addition, boron has other drawbacks, such as an ability to form deep centers like the D-center [117] rather than shallow acceptor states and, as shown in Section 4.3.2, boron ions also diffuse easily at the annealing temperatures needed for activation. The diffusion properties may be used in a beneficial way, although it is normally more convenient if the implanted ion distribution is determined by the implant conditions alone. 

Figure 4.27 Measured SIMS and SSRM current profiles of a multiple-energy implanted 4H-SiC sample. The implanted dopant ion is Al and the activation after two annealing temperatures is shown. (From [122], 2003 Material Science and Engineering B. Reprinted with permission.)...

Figure 18 Free carrier (electron) concentration in SiC samples implanted with P ions at RT (circles) and 1200 ° C (squares) as a function of annealing temperature. Annealing was performed for 20 min in Ar atmosphere. The electron concentration was obtained from Hall effect measurement at RT.

Figure 23. Examples of measured profile broadening of high-energy B implants at low annealing temperatures. SIMS data are from Ingram et al. (62). (Reproduced with permission from reference 59. Copyright 1988 Institute of Electrical and Electronics Engineers, Inc.)...

The fraction of the implant damage that anneals, fa, is assumed to depend directly on time at the annealing temperature ... 

Before discussing the redistribution of implanted dopants in GaN, it is necessary to briefly review the temperatures required to achieve electrical activity. Pearton et al reported that a temperature of 1100°C is required to achieve electrical activation of Si and Mg + P in GaN [3], However, this temperature is not sufficient to completely remove the implantation induced damage [4], To completely restore the crystal lattice, an annealing temperature of between 1250°C and 1600°C will be required [5], Most of the results on donor redistribution have been for anneals near 1100°C, as discussed in the following sections however, more experimental work must be done at the higher temperatures to characterise the effect of these higher temperatures. The following discussion is separated into common donor impurities (Si and O) and acceptor impurities (Be, Mg, Zn and Ca) in GaN. 

The Mg-implanted samples in FIGURE 1 do not show a conversion to p-type material for any annealing temperature and closely track the unimplanted material. However, the Mg-samples coimplanted with P show a conversion from n-type to p-type at 1050°C with an increase in hole concentration at the high temperature. The success in realising p-type material with the P co-implanted samples is attributed to the P acting to either (a) create additional Ga-vacancies, via additional implantation damage, for the Mg to occupy as acceptors, or (b) reduce the number of N-vacancies and thereby also increase the probability of Mg occupying the preferred Ga-sublattice [4],... 

FIGURE 1 Sheet resistance versus annealing temperature for unimplanted, Si-, Mg-, and Mg+P-implanted GaN (after [3]). 

---