The Distribution of Cl Impurities in SiO2 Films Produced by Thermal Oxidation of Si in Cl‐Containing Ambients

Abstract

C1 was introduced into SiO2 films by thermally oxidizing (100) Si in either OJHC1 or O2/C12 gas mixtures at llO0~ C1 concentration vs. depth profiles in these oxide films were determined by SIMS (secondary ion mass spectrometry). Analysis of the C1 distributions shows that, after incorporation into the SiO2 network at the interface, in the bulk of the growing oxide the bonded C1 is removed from the network. This process is analyzed quantitatively as a chemical reaction in which H20 reacts to replace C1 with OH. The kinetics of this replacement reaction and of the oxide film growth determine the C1 concentration vs. depth profiles in the grown films. The replacement reaction kinetics are faster when H20 is present than when only O2 is present. This accounts for the significantly different CI profiles in HCI oxides and Cl2 oxides. These observations are consistent with previous studies of the redistribution of implanted CI in oxide films during a second oxidation step.Oxidation of Si in chlorine-containing ambients is a widely used process in preparing gate oxide films for Si devices (1). In this process C1 is incorporated into the SIO2, and is found to be segregated primarily near the Si/SiO2 interface (2-4). Recent C1 concentration vs. depth profiles determined by secondary ion mass spectrometry (SIMS) show a significant amount of C1 distributed in the oxide with increasing concentration toward the Si/SiO2 interface (5-9). One interpretation of these profiles was that the measured C1 was determined by the high temperature distribution of a mobile, diffusing C1 species (7). The authors suggested that the C1 transport was by a "field-aided" diffusion process, against the concentration gradient. However, the origin of such a field was not established. In contrast, the C1 profile was suggested to be determined by a reaction-controlled process (9, 10). This difference has not been resolved. The question will be addressed in this paper.In a previous paper, we presented a semiquantitative analysis of C1 concentration profiles measured by SIMS (11). The profiles were obtained in the course of high temperature (900~ reoxidation of thermal oxide films implanted with C1; the analysis indicated that redistribution of the implanted C1 during the reoxidation step can be controlled by either reaction or diffusion. Which process dominates depends on the concentration of mobile molecular species (H~O or O2) capable of displacing chlorine species bonded into the SiO2 network (=-Si-C1) and on the relative magnitude of forward and reverse reaction rates in the displacement reaction. In fact, the redistribution process is slow and reaction limited in dry reoxidizing ambients (of order 1 ppm H20). It is fast and possibly diffusion limited in wet ambients (more than 1% H20).Similar processes should determine the distribution of chlorine within oxide films fabricated by thermal oxidation of silicon in ambients containing HC1. Because of the gas-phase reaction of HC1 and O2, the molecular species HC1, C12, 02, and H20 will be present in ...

The Effect of Fluorine Additions to the Oxidation of Silicon

“…As the oxidation temperature increases the fluorine continues to peak and saturate at a high concentration, broadening from this position at an almost constant concentration before decreasing to the outside interface. This shaped profile is directly analogous to the profiles found for chlorinated oxides (26). This behavior can be explained by a replacement reaction proposed for chlorine and the critical role of hydroxyl (OH) species in this mechanism.…”

“…The mechanism outlined in Ref. (26) needs to be pursued as a possible detailed description of these fluorine profiles. In contrast, the typical profiles obtained from the OjNF3 oxidation behave in a markedly different way as can be seen from Fig.…”

Direct Mass Spectrometric Identification of Silicon Oxychloride Compounds

“…The existence of a gas phase reaction also raises further questions regarding the origin of the Si3OC18 molecule. These results indicate that Si3OC18 may be formed by reaction of Si2OC16 with silicon tetrachloride Si2OC16(g) + SiC14(g) = Si3OC18(g) + C12(g) [2] or by an ion/molecule reaction between Si2OC15 § and SIC13 in the ionization chamber of the mass spectrometer Si2OCl~+(g) + SiC13(g) = Si3OC18+(g) [3] It has been noted that pure Si did not yield these volatile oxychlorides when the reactant gases were 1% C12/1% O2/Ar. In this case a SiQ layer limits attack of the Si substrate and very little SiC14(v) can form.…”

“…Thermodynamic calculations for silicon containing materials in a mixed oxygen-chlorine environment predict SiO2 as the stable phase; however, in regions of low oxygen potential, volatile silicon chlorides such as SIC14 and SIC13 may form (2). The formation of silicon oxychlorides has been suggested as well (3)(4)(5)(6)(7)(8)(9)(10)(11)(12). The purpose of this paper is to report some experiments which clearly show the existence of these species in the vapor form.…”

“…The previous work in this area centers on the incorporation of chlorine into the growing oxide on high purity silicon. Secondary ion mass spectrometry (SIMS) studies have shown that the chlorine tends to segregate at the Si/SiO2 interface (3)(4)(5)(6)(7). Monkowski and co-workers (7)(8)(9), have suggested that the chlorine actually reacts with the growing oxide to produce a silicon oxychloride phase.…”