Heinrich Koerner

Influence of barrier and cap layer deposition on the properties of capped and non-capped porous silicon oxide

Abstract Novel highly porous SiO 2 xerogels are being developed as low dielectric constant materials. For successful integration into DAMASCENE structures, the attractive electrical properties of these materials must not degrade as further cap and barrier layers are deposited and patterned. The influence of the deposition of PECVD SiO 2 cap and sputtered and MOCVD TiN barrier layers on the electrical properties of low k xerogel films was examined. FTIR was used to show that the pore surface methyl groups formed during HMDS treatment survive cap deposition. Electrical results indicate only small changes to the dielectric constant, leakage current density and field breakdown voltage after the cap was deposited. The deposition of the barrier layer was found to increase the dielectric constant of the xerogel by about 10–15% but not when the xerogel was capped first.

Stress induced metallurgical effects in Ti/TiN/AlCu/TiN metal stacks

Integrated circuits with aluminum metallization for products with high current densities need a metal stack with liner and antireflective coating (ARC) which can fulfill several requirements (e.g. low sheet resistance, high reliability, smooth surface, good adhesion, thermal stability, etc.). In this work different multilayer metal stacks are investigated and several phenomena which can be observed after thermal annealing of Ti/TiN/AlCu/TiN stacks are described and discussed. Metallurgical, electrical and mechanical properties of different layer combinations are investigated after thermal annealing and stress tests are done to compare the electromigration and life time behavior of each metal stack. For all investigated metal stacks it is shown that an interface reaction between Ti and aluminum will form TiAl3 phase. Even with very thick TiN layers on top of titanium or with only TiN liner the phase formation occurred. Explanations and models for the formation of different phenomena (hillocks, depressions ...

Growth kinetics of individual Al-Cu intermetallic compounds

The growth kinetics of Al-Cu intermetallic compounds (IMC) have been investigated on thin film couples and bonded samples in the range 150°C to 250°C using XRD, SEM/EDX and in-situ interface resistance monitoring. Individual diffusion constants D_o and activation energies E_a (1.01eV, 0.97eV, 1.23eV, 1.28eV) have been obtained from thin film couples for the main three IMC phases Al₄Cu₉, AlCu and Al₂Cu, and for the total IMC growth, respectively. Two additional phases (Al₃Cu₂, Al_0.06Cu_0.94) contribute to the total IMC growth at T ≥ 200°C, but do not form at lower temperatures. Lower activation energies of 1.13eV (thin film) and 1.05eV (bonded samples) have thus been derived for T <; 200°C for the overall IMC growth and are recommended to be used for lifetime predictions in the typical regime of device application temperatures.

Growth behavior and physical response of Al-Cu intermetallic compounds

This review covers recent investigations and concludes our findings for the growth of Cu/Al intermetallic compounds (IMC). [1, 2] The corresponding copper-aluminum interfaces were either established by a physical vapor deposited (PVD) Cu layer on a PVD aluminum pad or a Cu thermosonic nailhead bond on a PVD aluminum-based pad metallization. The identification, growth kinetics and mechanical strength of the different Al-Cu intermetallic compounds have been investigated. The annealing matrix of these investigations covered the temperature range from 150-300 °C for 25-2000 h. The identification of the Al-Cu phases utilizes X-ray diffraction analysis (XRD), selected area diffraction pattern (SAD) and scanning electron microscopy (SEM) & energy dispersive X-ray spectroscopy (EDX). The main three IMC phases Al4Cu9, AlCu and Al2Cu were identified over the whole temperature range, whereas two additional phases (Al3Cu2, Al6Cu94) contribute to the total IMC growth at temperatures above 200 °C. Individual diffusion constants D0 and activation energies Ea of 1.0 eV for Al4Cu9 and AlCu, 1.2 eV for Al2Cu and 1.3 eV for the total IMC growth have been obtained. As the two slow growing phases Al3Cu2 and Al6Cu94 were not observed below 200 °C, lower activation energies for the total IMC stack were expected and have been measured to be in the range of 1.05-1.1 eV for thin film and bonded samples for temperatures below 200° C. Therefore it is recommended to use these lower activation energies for lifetime predictions in the typical regime of device application temperatures. The impact of IMC thickness and annealing conditions on bond strength was studied using ball shear test. The test results did not show any hints on interface strength degradation across the full experimental matrix even for the groups where Al was already fully consumed, in case of a tungsten barrier or adhesion layer between Al metallization and silicon-based dielectrics was used.

Copper Contamination effect on the reliability of devices in the BiCMOS technology.

Copper (Cu) will be used to replace aluminum in the next generation metallization due to its low resistivity and high electromigration resistance. However, copper is a fast diffuser in silicon and silicon dioxide, and it is detrimental to the devices if it gets into the active region. We have investigated several approaches to contaminating with Cu the back surface of a fully processed BiCMOS wafer in order to study its effect on devices. In order to estimate the amount of Cu driven to the active region, a simulated drive-in diffusion experiment is used. Vapor Phase Decomposition--Atomic Absorption Spectrometry is used to measure Cu on the front surface of the wafer after annealing. In a fully processed BiCMOS wafer, the internal gettering: oxygen precipitation occurs at the initial high temperature process steps. This oxygen precipitation acts as trapping centers and an intrinsic barrier that prevents impurities that may be driven from the back surface of the wafer. The effectiveness of the internal gettering of a simulated BiCMOS processed wafer is measured in comparison to a monitor wafer which has no internal gettering. Electrical measurement shows an increase in the base current in a Gummel Plot measurement of the Bipolar device after Cu contamination. This effect is most visible for a wafer that has been annealed at 550 degree(s)C for 30 minutes.

Growth and reactivity of Al-Cu intermetallic compounds under ideal conditions

Cu wire bonding is fast coming up as Au wire replacement in semiconductor industry due to its lower price and several expected technological advantages. Cu bond wire reliability on Al pads is an ongoing study of interest and one of the mostly addressed topics is the degradation of Al-Cu intermetallic compounds (IMC) under high temperature or humidity stress conditions. The majority of published studies were carried out on Al-Cu systems in encapsulated packages. Many of them reported the degradation of the wire/pad-interface by cracks which are starting from the ball periphery and are penetrating inwards. Some also report about voids which are forming within IMCs or at interfaces. Questions have been raised about the root cause of the reported physical defects and the associated reliability risks. This paper is a more fundamental research investigating the Al and Cu diffusion reactions and the formation of individual IMC phases at clean and ideal Al/Cu interfaces without potential impacts of bond process and molding compound. It is conducted by using thin film couples of clean sputtered Al and Cu layers, which have been annealed under forming gas atmosphere at different temperatures (150°C-275°C) and for various time periods. Two different thin film couples have been used: a) 5000 nm Al followed by several μm copper to simulate an almost infinite Al reservoir and b) 650 nm Al followed by several μm Cu to simulate a typical Al reservoir after bonding. The samples were investigated by scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDX), Focused Ion Beam (FIB) and X-ray diffraction (XRD).