Dislocation Generation and Propagation during Flash Lamp Annealing

Abstract

Dislocation generation and propagation during flash lamp annealing for 20 ms was investigated using wafers with sawed, ground, and etched surfaces. Due to the thermal stress resulting from the temperature profiles generated by the flash pre-existing dislocations propagate into the wafer from both surfaces during flash lamp annealing. A dislocation free zone was observed around 700 μm depth below the surface of a 900 μm thick sawed wafer. The dislocation propagation can be well described by a three-dimensional mechanical model. It was further demonstrated that in wafers being initially free of dislocations no dislocations are generated during flash lamp annealing. © The Author During the last years, thermal processing on the millisecond scale was developed first of all for the creation of shallow dopant profiles in high end electronic device technologies and for the thermal treatment of layers.1,2 Flash lamp anneals (FLA) or laser anneals are able to heat the silicon wafer within a few milliseconds to temperatures up to the melting point. Depending on the irradiance of the flash, this can cause strong thermal gradients in the wafers inducing thermal stress which can generate defects. The aim of our investigations was to investigate dislocation generation and propagation during FLA. In addition, we modeled the temperature distribution and stress in a silicon wafer subjected to FLA and compared it to experimental results of dislocation propagation. The results of modeling help to understand the experimental results. ExperimentalThe experimental flow can be seen in Fig. 1. The wafers used for the experiments were 150 mm in diameter and of a resistivity of about 10 cm. The concentration of interstitial oxygen was in the range of 6.9-7.1 × 10 17 cm −3 (DIN 50438/1, new ASTM F121-83). In order to investigate dislocation generation and propagation during FLA, wafers of different surface state were prepared. Part of the wafers, 900 μm in thickness, was just sawed, another part was also ground down to a thickness of 750 μm, and the remaining part was also etched after grinding down to a thickness of 680 μm. After subjecting these wafers to FLA for 20 ms in nitrogen, (110) cleavage planes perpendicular to the surface were etched by Secco etchant for 3 min for dislocation delineation. The FLAs were carried out for 20 ms with a normalized irradiance of 0.97 using the flash annealing tool of the Institute for Ion Beam Physics and Materials Research of the HZDR. A tool for flash annealing usually contains both, halogen lamps for pre-heating and xenon lamps for the flash. Descriptions of typical tools for flash lamp annealing can be found in Refs. 1 and 2. In our case, the wafer was pre-heated for about 2 min from the back side by halogen lamps up to a temperature of 800• C. This helps to reduce the thermal gradients, the wafer stress, and reduces the energy required by the flash to reach a certain temperature. Then, the flash was initiated by the xenon-lamps energized by discharging a capacitor unit. The wafer was placed betwe...

A review of thermal processing in the subsecond range: semiconductors and beyond

“…This compares to the thousands of articles published on bulk silicon topics over the last ∼50 years. G Kissinger from IHP Frankfurt collaborated with people from Siltronic AG (formerly Wacker Chemitronic) as a main player in the wafer business [67][68][69]. Typical problems such as oxygen precipitation, minority carrier lifetime, and dislocation generation and propagation were highlighted.…”

“…Suppressing oxygen precipitation leads in this manner also to higher minority carrier lifetimes [68]. Also, dislocation generation and propagation during FLA were investigated using wafers with sawed, ground, and etched surfaces [69]. The thermal stress due to the temperature profiles generated by the flash makes pre-existing dislocations propagate into the wafer from both surfaces.…”

About the influence of deposited nitride layers on oxide precipitation after RTA pre‐treatment

“…The basics of the mechanical model are comparable to the ones used in Ref. . We adjusted an initial membrane strain in a way that the layer stress after deposition at 590 °C and cooling down to room temperature corresponds to the measured layer stress for the 85 nm nitride layers being about 1500 MPa.…”