No abstract

The capability of copper electroplating to produce void free filling of sub-micron high aspect ratio features has made it the process of choice for copper interconnect formation. Several aspects of copper electrodeposition including the basic electrochemistry and electrochemical kinetics, mass transport phenomena, potential gradients in solution, electrolyte composition, and the influence of various organic additives have been studied for over 50 years. Much of this basic understanding can be applied to development of integrated circuit (IC) copper electroplating processes. Other aspects of copper electroplating are unique to IC applications. These include the interactions of very thin seed layers with the electroplating process, the basic "bottom-up" filling mechanism necessary for seam free filling, and the metallurgical properties of sub-micron scale deposits. The dependence of IC filling processes upon plating bath chemistry and polarization characteristics are discussed in this paper.

The impedance characteristics of a copper disk electrode/sulfuric acid-cupric sulfate interface were investigated and interpreted in terms of equivalent circuit models. A single reaction mechanism, proposed to correspond to the Cu 2+ o Cu § charge transfer, limited deposition rates over a wide current density range at a well-agitated interface. A rate constant of 1.1 • 10 -6 cm/s, a transfer coefficient of 0.5, and a double layer capacitance of 25 jxF/cm 2 were determined from impedance data corresponding to the rate-limiting step. Arrhenius plots yielded an activation energy of 66 kJ/M and a pre-exponential term of approximately 10 ' cm/s. The involvement of a freely diffusing cupric ion in the rate-limiting step was consistent with the observed data, but the possibility of adsorbed species involvement cannot be eliminated. A minor impedance feature, believed to correspond to a relatively fast surface diffusion limited Cu + ~ Cu(O) charge transfer, was observed at low ac frequencies. In quiescent solutions, the observation of a significant Warburg impedance demonstrated that deposition rates became limited by diffusion of cupric ion to the interface.

Signalized intersections on high-volume arterials are often congested during peak periods, causing a decrease in through movement efficiency on the arterial. Much of the vehicle delay incurred at conventional arterial intersections is caused by high left-turn demand. Unconventional intersection designs attempt to reduce intersection delay and travel times by rerouting left turns away from the main intersection. Seven unconventional designs—the quadrant roadway intersection, median U-turn, superstreet median, bowtie, jughandle, split intersection, and continuous flow intersection designs—that could apply to a wide range of standard, four-leg intersections are compared. Previous comparisons of intersection delay and travel time between conventional designs and these unconventional designs have been piecemeal and have largely used hypothetical volumes. Simulation experiments were conducted using turning movement data from seven existing intersections of varying sizes to compare the travel time of conventional and unconventional designs fairly. Optimum cycle lengths were used for each design, and a number of factors were held constant to keep the comparisons fair. Off-peak, peak, and peak-plus-15-percent volume levels were examined. The results from the simulations showed that at each intersection one or more unconventional designs had lower total travel times than the conventional design. Whereas most of the unconventional designs showed improvement in one or more scenarios, the quadrant roadway intersection and the median U-turn designs consistently produced the lowest travel times. When considering the design of high-volume intersections like those tested, engineers should seriously consider quadrant roadway intersection and median U-turn designs where rights-of-way are available.

Conventional damascene electroplating uses a combination of organic additives, namely, a suppressor, an accelerator, and a leveler, to achieve superconformal fill of interconnects. This work demonstrates an alternative mechanism that produces bottom-up cobalt deposition through a combination of pH and suppressor gradient formation within the patterned features. The fill mechanism was investigated using voltammetric and electrochemical quartz crystal microbalance measurements. The results show that local pH affects both the deposition rate and the current efficiency for cobalt deposition, which, combined with the kinetic effects of suppressor-type additives, drive a plating rate differential between the field and the feature-bottom. By appropriately selecting solution concentrations, organic additives, the waveform, and the mass transport conditions, void-free superconformal cobalt fill can be achieved in a variety of features.