L.A. Larson

Thermal-wave measurements of ion implanted silicon

Abstract Results are presented of studies made to qualify photothermal-reflection spectroscopy for use as an ion implant process monitoring tool in a semiconductor production environment. The spectrometer used was the Therma-Probe 150 from Therma-Wave, Inc. Data is presented on the sensitivity of the system to the basic implant parameters: dose (fluency), energy, and species. Data is also presented on the sensitivity to ion channeling, dose rate, and time of measurement. The studies were limited to the low-to-medium-dose range, 1 × 1011 to 1 × 1014 ions/cm2, and to the energy range of 25 to 200 keV. Possible implications of the results on characteristics of implant damage are briefly discussed.

Damage formed by ion implantation in silicon evaluated by displaced atom density and thermal wave signal

Damage formed by BF+2 and As+ implantations in Si was evaluated quantitatively. The density of displaced atoms (Dda) was determined from 1.5 MeV He+ Rutherford backscattering spectrometery. Dda increased from 4.7×1016 to 1.6×1017 cm−2 with the dose increased from 6.0×1013 to 1.3×1014 cm−2. However, Dda saturates at around 4×1017 cm−2 for all doses above 5×1014 cm−2. The thermal wave signal intensity shows the same dose dependence as Dda. This result shows that thermal wave signal intensity has a close relation with the density of displaced atoms formed by ion implantation. Therefore, quantitative damage monitorings can be achieved by thermal wave intensity measurements. Also, the variation of thermal wave signal intensity with ion implant energy was studied.

Custom profiles by automated multi-step implantation

Abstract Multi-step implants using F as a preamorptuzing species for the formation of shallow p + source/drain junctions are investigated. SIMS, RBS and SRP data are combined to optimize process equivalents to 50 and 25 keV BF 2 implants. Incorporation of a dual implant sequence into a single, chained recipe and device results for a 1 Mbit SRAM are discussed.

Monitoring of low-dose ion implantation in silicon

Monitoring of low-dose arsenic or boron ion implantation (doses: 5*10/sup 10/ to 1*10/sup 13/ cm/sup -2/) in silicon, which is required for threshold voltage control of MOS transistors, is studied. The thermal-wave (TW) signal intensity decreases monotonically with decreasing dose. The lowest detection limit for As/sup +/ and B/sup +/ implantations is 5*10/sup 10/ and 1*10/sup 11/ cm/sup -2/, respectively. Correlation of the TW signal intensity versus damage density, TW intensity versus dose, and laser Raman intensity versus dose is obtained. The TW intensity is also correlated with the sheet conductance, and the threshold voltage of the transistor. Therefore, this technique is useful as a nondestructive, highly sensitive dose monitor for low-dose implantation to achieve tight threshold voltage control in MOS transistors.<<ETX>>

Dosing accuracy: A comparison of several experiments

Abstract Dosing accuracy is generally considered a moot point by the process engineer who is more concerned with obtaining good reproducibility. This situation changes radically when processes are transferred between fabs or when outside services are utilized. There appear to be no standards for what an implanted dose should be. Tests carried out at National indicate that dose will vary by machine type. The results of round robin experiments indicate that dose may vary by as much as +/- 20% partially by machine type but also through user practices. These experiments are compared and discussed along with several experiments attempted to compare results between separate tests.

An FTIR And RGA Study Of The Outgassing Of Photoresist During Ion Implant

Fourier Transform Infra-Red Spectroscopy (FTIR) along with total and partial pressure measurements have been used to characterize the severe resist outgassing that occurs during high current ion implantation. This outgassing is confirmed to be mainly the evolution of hydrogen as shown by Residual Gas Analysis(RGA). Hydrogen is liberated by the breaking of C-H bonds present in the polymeric component of the resist, thus resulting in carbon layer formation. This effect is clearly observed by FTIR. Due to this mechanism, outgassing is found to be independent of the heat treatments used to cure the resists for temperatures above that necessary to evolve solvents (90C). This study concentrates on high-dose, high current conditions. The outgassing behaviorfor these cases is saturated as a function of energy, current, and dose and the results indicate near total conversion of the resist to carbon.

Damage and RTA kinetics in Ar+ and Si+ ion-implanted CZ silicon characterized by thermal wave modulated optical reflectance

In this study, damage induced by Ar+ and Si+ ion implantation and its annealing behavior during rapid thermal annealing for 10 sec at temperatures between 575-1100°C were investigated by thermal wave modulated optical reflectance, deep level transient spectroscopy, reflection high energy electron diffraction, Rutherford backscattering aligned spectra and transmission electron microscopy. Our data show that (1) thermal wave signal and its variation with repect to rapid thermal anneal temperature strongly depend upon implant dose and anneal temperature; (2) both implant species induce four distinctive deep trap levels; (3) these traps evolve during rapid thermal annealing!; and (4) for the single Si+ ion implanted samples, the variation of total trap concentration with respect to rapid thermal anneal temperatures follows that of TW. However, in the case of Ar+ ion implanted samples, no apparent correlation between thermal wave signal and DLTS trap condition could be made.

The history of uniformity mapping in ion implantation

Abstract In the early days of semiconductor manufacturing, the four-point probe became established as the tool-of-choice for monitoring diffusion processes. The application of the four-point probe to ion implantation in the early 1960s was basically limited to the single-point measurement of dose, since the equipment did not have the necessary precision or repeatability to provide useful uniformity results. As a result, implanter uniformity was determined by either a visual observation of a heavily doped wafer or by a mylar bum. Unfortunately, these techniques were either subject to interpretation by the user or could not provide a parameter that could be statistically tracked or characterized. During the last 30 years, the commercial ion implanter as well as the dose and uniformity characterization equipment used to characterize this production tool have progressed significantly. Indeed, several generations of electrical and optical equipment have been developed to measure both the dose and uniformity of the increasingly advanced and complex ion implanter. This paper will review all of the techniques and equipment used to measure the uniformity of ion implantation. In addition, various graphical techniques developed to present early measurement results will be discussed.

A survey of implant particulate process control and yield effects

Abstract This paper is a review of current work to improve particulate control in ion implantation equipment. The first section includes current published research work which will be outlined and summarized. The second section focusses on work performed at National Semiconductor Corporation on testing and implementing an in situ particle monitoring system. The first equipment test consisted of a single monitor mounted in a medium current implanter. This test indicated that the particle signal correlated to surface scan monitor tests and that operating characteristics of the ion implanters which produce particles were observable. Our followup effort was an implementation of the in situ counters on all the implanters as a single particle counting system for the fab. The result of this work was a pareto-like analysis of machine and process issues which result in particle events. A correlation of lot-specific particle counts to yield was also developed. The advances in machine particulate control are contrasted with the needs and trends in process development. Although the advances in particulate control have been excellent, the predicted future requirements are even more stringent. The implications of these needs on both particulate performance and measurement are discussed.

The use of tertiarybutylarsine (TBA) and tertiarybutylphosphine (TBP) as liquid reagents for ion implantation

Abstract The safety and environmental concerns regarding the widespread use of phosphine and arsine gas in the semiconductor industry have given rise to the use of alternative, less hazardous replacement substances, such as tertiarybutylphosphine (TBP) and tertiarybutylarsine (TBA). TBP and TBA are less hazardous than the hydride gases under sudden release scenarios because they are (1) liquids, and (2) have a lower acute toxicity than the gases. For applications in ion implantation, the use of TBP and TBA could potentially reduce the costly downtime resulting from the heating and cooling cycles of the solid source vaporizers. In comparison with the gases, the use of liquids can extend the period of implanter operation before chemical sources need replacement. In this paper, data is presented which demonstrates the performance of these materials in a production environment, with the discussion focusing on issues such as source lifetime, implanter maintainability and overall performance.