TY - JOUR
T1 - Studying the effect of temperature on the copper oxidation process using hydrogen peroxide for use in multi-step chemical mechanical planarization models
AU - DeNardis, D.
AU - Rosales-Yeomans, D.
AU - Borucki, L.
AU - Philipossian, A.
N1 - Funding Information: This work was supported by the NSF/SRC Center for Environmentally Benign Semiconductor Manufacturing . The authors would like to thank Praxair, Inc. for gas cylinder donations.
PY - 2010/5/3
Y1 - 2010/5/3
N2 - Given that the mechanism responsible for removal during copper chemical mechanical planarization (CMP) is generally accepted to be based on the cyclic oxidation of copper and the subsequent removal of copper oxide, this study characterizes the copper oxide growth process as a function of temperature in aqueous hydrogen peroxide solutions. A copper oxidation model was proposed based on cation migration to adequately represent measured copper oxide growth profiles as a function of temperature. The two parameters extracted to fit the oxidation profiles, W and V, in the proposed model are related to activation energy of cation migration and the potential developed across the oxide film, respectively. The potential was found to be 0.95 V and did not vary with temperature. The activation energy was found to be 0.84 ± 0.01 eV and increased slightly with temperature. This slight increase, on the order of 2 to 3 kcal, has been previously reported and attributed to an increase in activation energy of cation solution in the oxide. The oxidized copper formation rates calculated suggest that the typical oxide thicknesses involved during the cyclic oxide growth and removal mechanism in copper CMP are between 7 and 12 Å. Though the oxidation model parameters are extracted from copper oxidation experiments on the minute time scale, there are a number of experimental, physical, and theoretical arguments that suggest the model represents the actual physical system and is applicable to the sub-second timescales involved during the oxidation processes in copper CMP.
AB - Given that the mechanism responsible for removal during copper chemical mechanical planarization (CMP) is generally accepted to be based on the cyclic oxidation of copper and the subsequent removal of copper oxide, this study characterizes the copper oxide growth process as a function of temperature in aqueous hydrogen peroxide solutions. A copper oxidation model was proposed based on cation migration to adequately represent measured copper oxide growth profiles as a function of temperature. The two parameters extracted to fit the oxidation profiles, W and V, in the proposed model are related to activation energy of cation migration and the potential developed across the oxide film, respectively. The potential was found to be 0.95 V and did not vary with temperature. The activation energy was found to be 0.84 ± 0.01 eV and increased slightly with temperature. This slight increase, on the order of 2 to 3 kcal, has been previously reported and attributed to an increase in activation energy of cation solution in the oxide. The oxidized copper formation rates calculated suggest that the typical oxide thicknesses involved during the cyclic oxide growth and removal mechanism in copper CMP are between 7 and 12 Å. Though the oxidation model parameters are extracted from copper oxidation experiments on the minute time scale, there are a number of experimental, physical, and theoretical arguments that suggest the model represents the actual physical system and is applicable to the sub-second timescales involved during the oxidation processes in copper CMP.
KW - Chemical mechanical planarization
KW - Copper oxidation
KW - Modeling
UR - http://www.scopus.com/inward/record.url?scp=77950538919&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=77950538919&partnerID=8YFLogxK
U2 - 10.1016/j.tsf.2009.12.089
DO - 10.1016/j.tsf.2009.12.089
M3 - Article
SN - 0040-6090
VL - 518
SP - 3903
EP - 3909
JO - Thin Solid Films
JF - Thin Solid Films
IS - 14
ER -