An electrochemical investigation of the removal and planarization of copper and ruthenium

With the continual scaling of integrated circuits past the 45 nm node, device limitations and the introduction of new materials are causing the industry to uncover new techniques and materials to improve the inner workings of electronic devices. Currently, chemical mechanical polishing (CMP) is considered one of the main challenges for chip manufacturing, due to the introduction of new low-K dielectric materials and high operating and consumables costs. Alternative CMP chemistries and methods are being investigated to keep up with ongoing scaling. In the present work, electrochemical mechanical planarization (ECMP) was investigated for use as a replacement or complement to Cu CMP. An electrolyte screening method was developed and tested. An optimal electrolyte was chosen as 1 M potassium phosphate with a BTA concentration of 0.001 M BTA with a pH value of ca. 2 at an operating potential of 0.5 V vs. Ag/AgCl reference.;Feature scale planarization has been shown to be particularly challenging when addressing the requirements of polishing technologies. Therefore, using the optimal electrolyte chosen, an ECMP tool was constructed in-house and used to test blanket and patterned wafer fragments. Individual features on the length scales of one to ten microm were fabricated through a series of plating and polishing regimes. These structures were used to study the feasibility of ECMP technology. Results using three pads types concluded that ECMP had the ability to planarize the Cu surface using the optimal operating conditions.;Ruthenium is being considered as a potential replacement liner material for tantalum/tantalum nitride (Ta/TaN) to prevent detrimental Cu migration into dielectrics. Various electrolytes were considered for their potential removal capabilities. Experiments concluded that acetate based electrolytes with a pH value of ca. 6.8 exhibited the best Ru removal compared to that of Cu. Additionally, the relationship between Cu and Ru is important due to the high probability of galvanic corrosion of the dissimilar metals usually under chemically harsh polishing environments. Due to the nobility of Ru compared to Cu, under certain circumstances Cu will be corroded in the presence of Ru leading to the decrease of overall production yields, the shortening of operational life-time and decreased device reliability. Galvanic corrosion between Ru and Cu was investigated using tafel plots and visual observation of bimetallic corrosion using an optical microscope, where two types of experiments were performed. Results concluded that developing an electrolyte screening method to predict the rate of galvanic corrosion through the use of corrosion potentials was not accurate. Also, it was demonstrated that the removal rate of the anode (Cu) was accelerated by initially exposing more area of the cathode (Ru) before experiments began.