A. Lietoila

Temperature distributions produced in semiconductors by a scanning elliptical or circular cw laser beam

Temperature profiles induced by a cw laser beam in a semiconductor are calculated. The calculation is done for an elliptical scanning beam and covers a wide range of experimental conditions. The limiting case of a circular beam is also studied. This calculation is developed in the particular cases of silicon and gallium arsenide, where the temperature dependence of the thermal conductivity has been taken into consideration. Using a cylindrical lens to produce an elliptical beam with an aspect ratio of 20, a 1‐mm‐wide area of an ion‐implanted silicon wafer was annealed in a single scan. The experimental data are consistent with the extrapolation of solid‐phase epitaxial regrowth rates to the calculated laser‐induced temperatures.

Epitaxial regrowth of intrinsic, 31P‐doped and compensated (31P+11B‐doped) amorphous Si

The rate of solid‐phase epitaxial regrowth of implantation amorphized 〈100〉 Si was studied in intrinsic, phosphorus‐doped and compensated (boron‐ and phosphorus‐doped) materials. The anneals were performed in flowing Ar gas in the temperature range from 477 to 576 °C, and the regrowth was analyzed by 2.2‐MeV 4He+ channeling techniques. The intrinsic and compensated samples exhibited nearly equal growth rates with thermal‐activation energies of 2.85 eV (intrinsic) and 2.8 eV (compensated). The growth rate in the 31P‐doped (constant concentration of 1.7×1020 cm−3) was enhanced by a factor of 6 to 8, while little change in the activation energy was observed.

Computer modeling of the temperature rise and carrier concentration induced in silicon by nanosecond laser pulses

A set of simultaneous equations for lattice temperature, carrier concentration, and carrier temperature is numerically solved for typical nanosecond laser pulses. The temperature dependences of the thermal conductivity and lattice absorption are included, as well as the free carrier absorption and reflection. Carrier diffusion and electronic heat conduction are taken into account, and Auger recombination is assumed to be the dominant recombination mechanism. The calculations show that while free carrier absorption plays a major role in annealing with 1.06‐μm radiation, only lattice absorption is important at wavelengths corresponding to photon energies well above the band gap. The Auger recombination coefficient is not a sensitive parameter, and the energy relaxation time does not affect the annealing results unless it is comparable to the pulse length. The results of the calculations are consistent with the hypothesis that the observed increase in silicon reflectivity is due to surface melting of the mat...

Calculation of carrier and lattice temperatures induced in Si by picosecond laser pulses

A previously presented computer model was used to calculate melt thresholds and carrier temperatures in crystalline silicon subjected to 20‐ps laser pulses at 532 nm. The energy relaxation time of hot carriers was a variable parameter. A thermalization time of 1 ps yields results which are in very satisfactory agreement with published experimental data: the melt threshold is 0.19 J/cm2 and the maximum carrier temperature for the threshold pulse is 5000 K.

Metastable As‐concentrations in Si achieved by ion implantation and rapid thermal annealing

Rapid thermal annealing (4 min at 560 °C) of implanted As in Si can electrically activate concentrations up to 5×1020 cm−3, which exceeds the solubility at that temperature. Prolonged thermal annealing at the same temperature causes deactivation of such metastable concentrations. The results suggest that when the amorphous‐crystalline interface is moving during solid phase epitaxial regrowth, dopants go to substitutional sites even in concentrations exceeding solubility. If solubility is exceeded, deactivation takes place after the material has recrystallized.

Temperature rise induced in Si by continuous xenon arc lamp radiation

It is shown that practical beam annealing of silicon can be accomplished with a high intensity arc lamp. The use of a one‐dimensional, steady‐state solution for temperature is justified. The Kirchhoff transform is utilized to include the temperature dependence of the thermal conductivity. Surface temperatures produced by a xenon arc lamp are calculated for 300‐ and 375‐μm thick silicon samples, using substrate temperatures of 350 and 500 °C. It is shown that substantial reduction of the radiation intensity required for a given surface temperature can be obtained by placing a quartz wafer between the silicon sample and the heat sink.

Photovoltaic measurement of bandgap narrowing in moderately doped silicon

Abstract Solar cells have been fabricated on n-type and p-type moderately doped Si. The shrinkage of the Si bandgap has been obtained by measuring the internal quantum efficiency in the near infrared spectrum (hv = 1.00−1.25 eV) around the fundamental absorption edge. The results agree with previous optical measurements of bandgap narrowing in Si. It is postulated that this optically-determined bandgap narrowing is the rigid shrinkage of the forbidden gap due to many-body effects. The “device bandgap narrowing” obtained by measuring the pn product in bipolar devices leads to discrepant values because (i) the density of states in the conduction and valence band is modified due to the potential fluctuations originated in the variations in local impurity density, and (ii) the influence of Fermi-Dirac statistics.

Deep levels in ion‐implanted, CW laser‐annealed silicon

Electronic defect levels in self‐implanted, CW Ar‐laser‐annealing silicon have been measured by deep‐level transient spectroscopy. Results are presented from measurements on n‐type Schottky diodes which were fabricated on both Czochralskigrown an epitaxial silicon wafers. High densities of electron traps remain after CW laser‐induced recrystallization of an implanted amorphous layer. The defect densities decrease with depth into the silicon substrate. The residual damage is only partially decrease with depth into the silicon substrate. The residual damage is only partially removed by a 600‐C anneal and is substantially removed after an 800‐C anneal.

Surface Temperatures Produced In Silicon Using Large Diameter Scanning Cw Sources

The temperature induced at the surface of a silicon wafer is calculated for both circular and rectangular scanning beams, assuming both a thermally isolated and heat sunk back surface. In the former case, the problem is one of simple uniform heating with radiative cooling at slow scan speeds. A significant thermal gradient is produced across the wafer at high scan speeds. In the latter case, a one-dimensional analysis suffices except for beam diameters less than about twice the wafer thickness, when lateral heat diffusion effects are present.

Chapter 3 Applications of CW Beam Processing to Ion Implanted Crystalline Silicon

Publisher Summary This chapter discusses the analytical techniques used in the evaluation of CW laser and e-beam annealing of ion implanted single crystal silicon. Solid phase epitaxy is discussed and the results of CW beam processing in amorphous and nonamorphous Si are reviewed. The crystalline quality, as judged by transmission electron microscopy (TEM), is good in CW laser annealed silicon layers. Results obtained using a scanning e-beam, both in an electron beam welder and in a scanning electron microscope, are largely similar to those for CW laser scanning. The most significant difference between these methods and the use of a laser occurs if the silicon sample to be annealed is covered by an oxide or other dielectric layer. If a laser is used, the dielectric layer will act as an antireflective coating, and the laser power must then be adjusted accordingly. This may cause problems in device structures, where the oxide thickness is not uniform throughout the scanned area. For the e-beam, however, this problem does not exist, because no reflection takes place either from the silicon or from silicon dioxide. On the other hand, charging of the oxide may be important for certain choices of the beam current and voltage.