John K. Lowell

Randomly oriented Angstrom‐scale microroughness at the Si(100)/SiO2 interface probed by optical second harmonic generation

Femtosecond pulses from a Kerr–Lens mode‐locked Ti:sapphire laser are used to generate second harmonic from a series of native‐oxidized Si(100)/SiO2 and hydrogen‐terminated Si(100) samples prepared with systematically varied interfacial microroughness with root‐mean‐square feature heights ranging from 0.6 to 4.3 A. Rotationally anisotropic second harmonic signals using different polarization configurations were measured in air and correlated with atomic force microscopy measurements. The results demonstrate rapid, noncontact, noninvasive measurement of Angstrom‐level Si(100)/SiO2 interface roughness by optical second harmonic generation.

Electrochemical Investigation of Copper Contamination on Silicon Wafers from HF Solutions

Copper contamination of silicon wafers from 50 :1 HF solutions containing 0 to 100 ppb Cu was studied using dc electrochemical techniques. As the level of copper concentration in HF solutions increased, the corrosion current density and corrosion potential of silicon as well as the amount of copper deposition were increased. Upon addition of a nonionic surfactant, the corrosion potential, corrosion current density, and the extent of copper deposition were decreased. However, the levels of deposited copper and surface roughness were dependent on surfactant concentration. When H 2 O 2 was added to copper-spiked HF solutions, the open-circuit potential of silicon recovered to a value that is characteristic for silicon immersed in a mixture of H 2 O 2 and HF indicating the removal of deposited copper on silicon.

Surface barrier detection in plasma‐enhanced chemical vapor deposition oxides

The performance standards of plasma‐enhanced chemical vapor deposition (PECVD) oxides increase as device sizes shrink and densities increase for ULSI. Future PECVD dielectric quality may be compromised by the effects of contaminants and/or charged ionic species that are present in the gaseous ambient which the wafers are exposed to. As a result of being left at the oxide/silicon interface, these gaseous ambient defects such as metallic ions can adversely effect reliability and performance. Moreover different types of PECVD oxides require different deposition chemistries and are grown to differing thicknesses making it difficult to apply a single C–V test. In this article, the problem is addressed by detecting, nondestructively, such residual interfacial charge postprocessing on product wafers rather than the traditional test wafers. For the first time the effects of mobile ions and interface states at the oxide/silicon interface and in the as‐processed films for various PECVD oxides will be shown using no electrical test methods before and after anneal. Charge effects and approximate generation lifetime estimates are separated out from the total charge measurement using passive high‐injection surface photovoltage. Thus one can passively examine the induced surface barrier height and interface charge for any PECVD oxide type at any thickness. This technique is demonstrated for several types of single‐oxide samples. The charge effects (if any) in PECVD oxides as‐deposited with or without rapid thermal annealing treatments are illustrated.The performance standards of plasma‐enhanced chemical vapor deposition (PECVD) oxides increase as device sizes shrink and densities increase for ULSI. Future PECVD dielectric quality may be compromised by the effects of contaminants and/or charged ionic species that are present in the gaseous ambient which the wafers are exposed to. As a result of being left at the oxide/silicon interface, these gaseous ambient defects such as metallic ions can adversely effect reliability and performance. Moreover different types of PECVD oxides require different deposition chemistries and are grown to differing thicknesses making it difficult to apply a single C–V test. In this article, the problem is addressed by detecting, nondestructively, such residual interfacial charge postprocessing on product wafers rather than the traditional test wafers. For the first time the effects of mobile ions and interface states at the oxide/silicon interface and in the as‐processed films for various PECVD oxides will be shown using no...

An Optical Approach to Evaluating the Effects of Chlorine on the Quality Of Si/SiO 2 Interfaces

In CMOS, the addition of chlorine particularly in TCA form to the growth of thermal oxides in logic technologies is well-known and pervasive. In addition to the increasing environmental concerns of chlorine use, one of the important parameters is the amount of metallic contamination due to transition metals such as Fe in the Si, and alkali metals like Na in the oxide since these phenomena effect both device performance and quality. However, the ability to measure this parameter on product material is not generally available due to inherent problems with most known methods. In this paper we report on the application of high-injection, frequency based optical surface photovoltage (SPV) and a more recent technique known as a contact potential difference (CPD) to both quantify and qualify as-grown oxides on CZ P-type silicon.

Evaluation of contaminant-induced charge from oxide chemical-mechanical polish

The role of chemical-mechanical polishing (CMP) of ILD and metal layers is increasing as device densities in the ULSI shrink. However, the advantages in film planarization gained with CMP are often offset by contaminants which are abundant in the chemical bath (known as the slurry) to which the wafers are exposed during processing. Such chemistries particularly metallic ions can adversely effect device reliability and performance if left on the wafer or ILD in large densitites. The problem we address is that of detecting such residual contaminants post-CMP on product wafers nondestructively. In this work we look at the effects of ionic residuals in ILD oxides before and after exposure to various methods of CMP. Using an optical methodology known as contact potential differentiation in which the potential across the oxide is separated out from (and compared with) a standard surface barrier measurement, we can passively examine any dramatic charge on the wafer and in the oxide as a result of this process. In this paper this technique will be demonstrated on several CMP samples illustrating the contaminant effects on CMP oxides with results compared to chemical spectroscopy.

Influence of oxygen-iron interaction on the external gettering of Fe in p-Si by polycrystalline silicon film

External gettering of iron by thin polycrystalline silicon film in p-type CA silicon has been investigated using deep level transient spectroscopy, and the surface photovoltage method. Depth profiles of iron concentration indicated a sharp gradient in the Fe concentration near the polycrystalline silicon/substrate interface. Concurrent decrease in the micority carrier diffusion length was also observed in the same region. The majority of iron gettering from the bulk silicon was found to be associated with the enhancement of the internal gettering. The presence of small oxygen precipitates/nuclei generated by prolonged heat treatment in the range of 600C-700C was found to prevent regeneration of FeB pairs at the room temperature. Similarily, carbon in the bulk silicon was found to retard the regeneration of the pairs. On the other hand, large precipitates formed at 1000C do not influence the diffusion or the recombination of Fei with B- to form FeB pairs.

Ex-situ and in-situ probing of Column IV interfaces using optical second harmonic generation

Using femtosecond pulses from a Kerr-lens Mode-Locked Ti:Sapphire laser to generate second harmonic from a series of native-oxidized Si(001)/SiO2 samples prepared with systematically varied etch- induced interfacial microroughness, we demonstrate rapid, noncontact, noninvasive measurement of Angstrom-level Si(001)/SiO2 interface roughness. These measurements were performed in air and correlated with atomic force microscopy (AFM) measurements. We also demonstrate in-situ second harmonic monitoring of Si epitaxy in two growth regimes: high temperature (approximately equals 925 K) ultra high-vacuum chemical vapor deposition (UHV-CVD) growth mode and a cyclic atomic layer epitaxy (ALE) growth mode. During UHV-CVD growth of Si on ALE-grown Si0.9Ge0.1(001), we observed interference of the second harmonic signals between the growing Si surface and the buried Si0.9Ge0.1(001) interface. In the ALE growth mode, we monitored several key stages during a full cycle of growth of a partial (approximately equals 0.42) Si monolayer on Si(001) from a disilane (Si2H6) precursor.

Characterization of surface barrier effects from Cu-implanted SiO/sub 2//Si interfaces

The effects of Cu in silicon as a contaminant ion have been studied by many groups. As a result its effect on gate oxide integrity through surface defect decoration and retardation of thermal oxide growth have become recognized problems for VLSI and ULSI. However its effects on surface charge, barrier height, and minority carrier generation are now as well-understood since the ability to study Cu-induced surface defects directly from surface cleans or implant processing is reduced by the addition of other contaminant species or the compromising effects from additional process steps required to fabricate C-V test capacitors, In this study we address this problem by using a frequency-swept, high-injection surface photovoltage method which follows the direct, ex-situ examination of the surface barrier condition and interfacial states affecting generation lifetime. For this investigation we have generated a variety of Cu-contaminated samples fabricated by ion implantation over three oxide thicknesses and two Cu surface concentrations. These are compared to samples with Cu introduced by buffered-oxide-etch (BOE) and electrical results from standard C-V tests. By using passive optical probes, we present direct observation of surface state conditions due to the presence of Cu.

Non-destructive detection of ion implant contamination: a SEMATECH/AMD study

In this paper we report on a systematic study of ion implantation equipment currently in operation or development by IC manufacturers and equipment vendors. The application of optical surface photovoltage (SPV) to both quantify and qualify bulk implant-induced contaminants in CZ P-type silicon is emphasized. We address the issue of contaminants and exemplify the use of SPV as a passive, in-line technique for assessment of the problem.

Oxide charging induced by electron exposure in ion implant

The use of electron showers (flood guns) during ion implantation has been used industry-wide to compensate the parasitic effects of electrostatic charging induced by energetic ion beams impinging on the surface during processing. Moreoever for damage protection the wafer may be covered by a protective oxide which inhibits the majority of secondary electrons in the exposure path from reaching the wafer surface. Previous work has shown that the surface barrier of the oxidized surface does not change significantly when flood guns are used which supports this premise. However, in this work we are interested in what effects the electron exposure may have on the oxide potential.