Handbook of Modern Coating Technologies

Rutherford backscattering and elastic recoil detection

Methods based on the registration of RBS energy spectra and recoil nuclei [elastic recoil detection analysis (ERDA)] formed in sliding interactions of beam ions with the surface are especially popular for ensuring high accuracy of the analysis of element depth distribution profiles. The depth resolution of the most common semiconducting detectors of charged particles is ~10 nm, compared with ~1 A for detection utilizing specialized spectrometers. In doing so, the samples must have a polished surface. The general physical principles of these methods are expounded in books [88,89]. Surface roughness remains the main factor hampering the achievement of high-resolution detection for commercial and natural samples and rendering interpretation of experimental data difficult. A new approach to parameteriz­ing Monte Carlo calculations is proposed in Ref. [90]. Numerical codes based on the analytic formulas describing physical processes behind beam particle scattering can be used to rap­idly calculate roughness-related effects for various uneven surfaces. Similar to PIXE, RBS spectrometry is supported by several computational programs permitting interpretation of backscattering spectrum profiles. The results of their independent testing are reported in paper [91].

By virtue of its kinematic characteristics, RBS spectrometry is most efficacious for the investigation of the local distribution of heavy elements in light matrices, for example, for determining heavy metal impurities in biological objects. ERDA is extensively applied to study hydrogen concentration profiles in near-surface layers of various materials [92].

• Nuclear reaction analysis

Nuclear reaction analysis (NRA) or analysis of instantaneous emission accompanying nuclear reactions has its origins in interactions between light ions with an energy of several megaelectron-volts with sample atoms. An ion can overcome the Coulomb barrier of an atomic nucleus and approach as close to it as the nucleus radius; it may induce a nuclear reaction resulting in a structural change of the nucleus. The products of such a reaction are hydrogen and helium ions, neutrons, and g-radiation recorded with detectors. The energy dependence of the nuclear reaction cross section has a few narrow resonances. Therefore the probability of a reaction is especially high when beam ions possess an energy of several megaelectron-volts.

As the beam energy grows and reaches a resonant value, nuclear reactions involving one kind of atoms begin to occur on the sample surface. A further rise in the beam energy causes a resonance reaction in deeper layers as a result of ion deceleration, which makes it possible to determine element depth distribution profiles with a resolution of ~10 nm and a sensitiv­ity of ~0.1 ppm [93]. Because the Coulomb forces for heavy nuclei markedly decrease the reaction cross section, interactions with the light nuclei of the sample, Z < 15, are most effi­cacious within a beam energy range <5 MeV. The high selectivity of NRA is due to different energy spectra and cross sections of nuclear reactions for different elements and their iso­topes. Experimental facilities for local microanalysis based on the NRA method with the use of SNMPs are described in Ref. [94].

• Registration of ion beam-induced charge

Registration of ion beam-induced charge (IBIC) found a wide application in the 1990s for studies of microelectronic devices, semiconductor radiation detectors, solar cells, distribution of dislocations, etc. This method relies to the full extent on peculiarities of the passage of light ions accelerated to MeV energies in semiconducting materials and insulators. Their small deflections from the straight trajectories ensure high spatial resolution of the method compared with electron beams. The method is based on the formation of electron-hole pairs due to ion energy transfer during the beam passage in a semiconducting material. The possi­bility of determining the number of newly formed electron 6 hole pairs is due to a number of internal and external factors, such as recombination at point and extensive defects, impu­rity concentrations, diffusion length of minority carriers, and electric field strength. The IBIC method utilizes focused light ion beams with an energy of several megaelectron-volts and weak currents (0.1 — 1 fA). It measures individual charge pulses. Analysis at such low beam currents is possible because each ion generates a sufficiently large number of electron-hole pairs as it passes through a semiconducting material or insulator and the net charge deter­mining the signal amplitude exceeds the noise level of the measuring instrument.

Fig. 5—8 depicts major types of measuring circuits used for registration of IBIC (see review [95] devoted to the general physical principles, the theoretical basis, and application areas of the method in question). It can be seen that charge carriers formed in the depleted region of p—n junctions slowly diffuse from the place of their generation;many of them recombine at point defects, which accounts for a limited number of carriers reaching the contacts of the instrument. The time interval during which a pulse is recorded varies from pico- to microseconds. Under normal measurement conditions, the charge-generated pulse is first accelerated to several volts. Thereafter, the digitized signal marked by the beam posi­tion is accumulated in computer memory by the data acquisition system built around the charge-coupled devices. The data processing yields information on the sample microstruc­ture in the form of a contrast image of the concentration of electrons and holes [96,97]. Apart from registration of the total accumulated charge in the case of a fixed beam position, it is just as well possible to measure induced charge evolution time from pulse variations at the contacts (time-resolved IBIC or TRIBIC method). This, in turn, permits estimating the mobility of charge carriers in the sample [98].