The use of ultraviolet (UV)/ozone treatment for the removal of hydrocarbon contamination is discussed for a number of substrates involving complex composition or geometries. Treatment of gold and silicon dioxide surfaces in a reactor attached on line to an x-ray photoelectron spectrometer showed that the chemistry of oxidized hydrocarbon on SiO2 differs from that on gold and the removal rate of hydrocarbon bonded to SiO2 is much faster than that on gold. Treatments of thin film transistor surfaces consisting of field oxides and metallurgical strips showed that considerably longer reaction times are required to clean the field oxide regions compared to treatment of a single component surface. Cross-contamination of specimen surfaces can occur if labile contaminants such as fluoride are present anywhere on the structure. UV/ozone treatment has also been investigated for obscured surfaces not directly irradiated by UV. Effective cleaning is still directly possible using ozone alone as a reactant, although the reaction time is longer.

No abstract

Tinplate surfaces have been analysed on a microscopic scale for differences in composition which are related to their effectiveness in providing cathodic protection of the underlying steel. Secondary ion mass spectrometric images of a number of elemental and molecular ions have been collected and collated in stacks that represent the three-dimensional distribution of the ions in the tinplate. An iron-tin alloy structure at the interface of the tin and steel appears to be detectable by the measurement of the 176FeSn-molecular ion image. The localized corrosion behaviour of the tinplates studied has been correlated to the microscopic distribution of iron near the outer surface of the tinplate; the presence of such an iron-rich phase could increase the dissolution rate of surrounding tin and thus reduce the effectiveness of the cathodic protection provided to the steel substrate. The steel substrate was shown to have a low but detectable concentration of oxygen from its interface with tin down to a depth of 1-2 pm. INTRODIJCTIONSecondary ion mass spectrometry (SIMS) has demonstrated unique strengths when applied to the investigation of metallurgical structures at or near the surface.'-3 The high dynamic range of the secondary ion signals imparts a capability to measure phase composition over a wide range of elements, including light elements such as hydrogen, carbon and oxygen. Combined with the excellent depth resolution (< 5 nm) and very low detection limits (ppm for most elements and ppb for some), such measurements of phase composition are doubly useful if they can be carried out within microscopic areas of any domain, since the metallurgical phase structure is usually quite complex at gas/solid and layer boundaries.The structure of tinplate on steel is a good example of such complexity. Between the outer surface of pure tin (usually 0.15-2 pm thick) and the steel substrate (0.1-0.4 mm thick), one or more iron-tin alloys are proposed to form, particularly after the electrodeposited tin is subjected to a reflow treatment.4 Depending on the tin coating thickness and the reflow treatment used, the thickness of the iron-tin alloy layer varies from 0.15 to 2 pm. In general, the pure tin layer acts as a sacrificial electrode, thus protecting the steel substrate from corrosion. Further protection to the steel substrate, to some extent, is accorded by the iron-tin alloy layer, which acts as a barrier to solution exposure.The quality of the tinplate and its resistance to localized corrosion, to some extent, depend on the steel substrate. For example, the thickness and the uniformity of tin coating are affected by the rolling process prior to electroplating. After electroplating, the tin coating profile closely follows the rolling marks on the steel substrate. On subsequent reflow treatment, molten tin spreads evenly on the substrate, resulting in thicker tin layers at the troughs and thinner tin layers at the crests

No abstract

With ongoing demand for high density wiring and high I/O on VLSI chips, the requirement of high wire bond yield is a challenge to achieve low cost, high performance and reliable products. Secondary Ion Mass Spectrometry (SIMS) was used to investigate the metallurgical contaminants on the gold wire bond pads and their impact on wire bond yields. SIMS depth profile studies showed that copper and nickel in concentrations greater than 1 wt% caused poor wire bondability, while copper concentration at less than 0.1 wt% resulted in good bondability of Al ultrasonic wire bonded to the gold pads.