Implantation Profiles

Ion implantation is a method commonly used for doping semiconductors. Because the concentrations of the dopants (mostly B and P) are very low, a dynamic range of more than five orders of magnitude is often necessary. Measurement of is more difficult than that of B, because of the mass interference of °Si H. High mass resolution of m/Am = 5000, or an energy offset of 300 V, is necessary. 

The depth profiles in Fig. 3.26 show that the typical flat channeling implantation profile is generated with low doses only. Increasing the dose superimposes the normal implantation profile shape. Undertaking such experiments with homogeneous wafers enables the production of calibrating models for semiconductor production. 

---

The physical techniques used in IC analysis all employ some type of primary analytical beam to irradiate a substrate and interact with the substrate s physical or chemical properties, producing a secondary effect that is measured and interpreted. The three most commonly used analytical beams are electron, ion, and photon x-ray beams. Each combination of primary irradiation and secondary effect defines a specific analytical technique. The IC substrate properties that are most frequendy analyzed include size, elemental and compositional identification, topology, morphology, lateral and depth resolution of surface features or implantation profiles, and film thickness and conformance. A summary of commonly used analytical techniques for VLSI technology can be found in Table 3. 

Random and channeled implanted profiles in 6H-SiC are shown in Figure 4.11(a). These profiles correspond to implants performed at RT with 1.5 MeV AF to a low fluence ( 10 cm" ). The flux is not accurately defined because of the experimental constraints, as explained in [74]. In fact, most of the studies concerning channeled implants make use of ion beams in single- spot configurations and have to face the problems related to nonhomogeneous flux distribution within the beam spot. The... 

Inserting values for D+ and r into the expression for L+, a value of approximately 1000 A is obtained, and this is typical for metals. One can then find an estimate of the efficiency e of the moderator by multiplying the implantation profile P(x), equation (1.9), by the probability that a positron reaches the surface from a depth x, exp(—x/L+), and integrating over all values of x. Then, since pimpL+ -C 1, the efficiency may be written as... 

Brandt, W. and Paulin, R. (1977). Positron implantation profile effects in solids. Phys. Rev. B 15 2511-2518. 

The helium implantation profiles, after incorporating the angular distribution for the 3.5 MeV alpha particles envisioned in a power reactor, are roughly Gaussian in shape and peaked at 1.5 to 4 pm. Calculated helium profiles show that surface deformation can occur if a critical concentration is reached. The likelihood of blistering is then determined by a competition between the accumulation of helium at a depth of several microns, and the erosion of the surface by D-T sputtering. 

An excellent way to create standards is ion implantation of the elements of interest into the matrix. This works exceptionally well for semiconductors since one can usually start with high-purity single-crystal materials that represent the matrix of interest. Also the use of Eq. (4.8) is well suited for this purpose since ion implanters usually quote doses in atoms per square centimeter. However, Eq. (4.5) serves just as well by converting the matrix concentration to atoms per cubic centimeter. In this procedure, the implant profile is sputtered through, the implant element secondary ions and the matrix element secondary ions are each summed, and the depth of the sputter profile is determined, usually by using a stylus profilome-ter. The sensitivity factor is then calculated from... 

FIGURE 2 compares the room temperature PL data from Zn-implanted GaN samples annealed at 1150°C and 1250°C under 10 kbar N2 with an MOVPE-doped GaN Zn reference sample and the undoped GaN/sapphire starting material [8], The MOVPE Zn dopant concentration was designed to closely match the implantation profile to provide an accurate comparison of the relative Zn luminescence efficiencies. The spectra observed in all of the Zn-doped samples are similar, differing rally in intensity, which confirms the activation of the implanted Zn acceptors. The sample annealed at 1250°C shows intense Zn PL roughly an order of magnitude stronger than in the epitaxial GaN Zn reference sample... 

The difference in the annihilation ratio for positronium at the surface and in the sample is used to measure the effective range of positronium. Implantation profiles for a range of incident energies (density 1 g/cm3) were calculated. In this simulation the fraction that stops within a diffusion length of the surface can reach it and annihilates into 3 photons the remainder annihilate 10% into three photons and 90% into two photons, as shown in Figure 7.4. The fractions are chosen as an example. A short effective range appears as a sharp transition from surface measurement to bulk measurement value. 

A simple fit of the data with the product of an exponential association and an exponential decay to estimate the escape depth, overestimates the escape depth by folding the positron implantation profile and diffusion into the fitting parameters [30], A more appropriate numerical fitting method based on the diffusion equation was used to take both the implantation profile and diffusion into account [31]. When it is applied to the 3-to-2 photon ratio data suitable absorbing boundary conditions need to be included. The results for the escape depth are shown in Figure 7.8 [30]. In addition to the diffusionlike motion of positronium in connected pores, positrons and positronium diffuse to the pores. 

The fit results from 3-to-2 photon annihilation data can also be used to determine the fraction of the total porosity, which is open to the surface. The fraction is determined by comparing the asymptotic values of the fit for the total signal and the constant part. Here, too the implantation profile and the motion of positrons and positronium have to be taken into account. The results are shown in Figure 7.9. 

In the first case, open symbols, a thin skin layer was created by an etch treatment. The second case shows a sample capped with a 200 nm thick cap. Either treatment seals the surface and prevents the escape of positronium. Below the surface layer, significant porosity can be observed. However, because the implantation profile is spread across larger depth regions in the second case, a careful data analysis must be undertaken to extract the actual porosity. 

The author would like to thank Dr. James R. Ehrsteln of the National Bureau of Standards for providing Figures 6a and 6b and Dr. Marek Pawllk of the GEC Research Laboratories for supplying the boron implant profiled in Figure 8. 

In subsequent experiments, Biersack, et al. (29) used the boron (n,alpha) reaction to show the effect of pre- and post-irradiation damage on boron implantation profiles. By post-irradiating a boron implant in silicon with 200 keV H2, a migration of the boron to the Induced damage sites was observed. In the same paper, diffusion and trapping of lithium ions in niobium were reported. Using the lithium (n,alpha) reaction, they mapped irradiation Induced crystal defects through a depth of several micrometers with respect to several sample treatment conditions. 

Boron profiles by NDP in cadmium mercury telluride, an important infrared detector material, have been measured by Ryssel, et al. (31) and Vodopyanov, et al. (32). Cervena, et al. ( ) used NDP to study the implantation profiles of °B in several photoresists used in masking operations and to determine range values for Implants in several types of grown or deposited SIO2 films. 

---