Efficient Determination of Copper Electroplating Chemistry Additives

Determination of copper electroplating additives is critical to ensuring consistent copper plating of conductors and through-silicon-vias used in semiconductor processing and electronics packaging. The present analysis methods require many chemical analysis steps, generate waste, and are not very accurate. A new set of analysis methods utilizes a reduced number of steps and along with the use artificial neural networks (NNs) overcomes the present limitations, and provides an accurate and ecologically green analysis procedure. Using a newly developed second-order NN algorithm called neuron-by-neuron analysis, accuracies less than 1% have been realized. A step-by-step procedure to implement this method is provided.