The Practical Method for Monitoring Additives in Copper Electroplating Baths Using the Chronopotentiometry Technique

1. Introduction
   We, Toppan Printing, developed the electrochemical method for monitoring copper sulfate electroplating bath ¹⁾and have been applying it to FC-BGA manufacturing.  We describe the outlines of the method and currently acquiring knowledge.

2. Electrochemical Measurements
   We use chronopotentiometry technique for practical monitoring.  The measurement is performed using setup with usual equipment, and time-potential curve is obtained.  The features of the curves are affected by the additive condition in the bath and able to reveal not only the composition variation but the degradation of additives, caused by long term operation.  By means of the time-potential curves, the bath characters can be identified.
   The typical results are shown in Fig.1-Fig.3.  The curves were obtained for three kinds of baths and the conditions before and after aging when copper anode was used.   The curves show the characters of the baths and how the baths vary by aging.  Moreover, it is found that, excepting very initial portion (time<100s or less), the most of curves show the behaviors of positive potential shifting as the time proceeds.  The positive shifting should indicate that the acceleration overcomes the inhibition on the electrode surface.

3. Analytical Methods
3-1 Numerical characterization
   The numerical analyses of the curves must be important to quantify the curves.  We divide the curves to some parts. It is found that the slope values at first part and the average values at last part are available to judge the bath conditions, whether via-filling is OK or NG, as shown in Fig.4.  In this case, the saturation potential value in the last part can be correlated with the via-filling properties rather than the slope in the first part.  The averaged saturation potential values (obtained in the range from 800s to 1200s) were -139, -141 mV vs.Ag/AgCl for via-filling OK, and -118 mV vs.Ag/AgCl for via-filling NG.   

3-2 Curve fitting with the equation based on the mechanisms
   Because the time-potential curves express the behaviors of inhibitors and accelerators functioning on the electrode surface, the parameters indicating the additive behaviors can be obtained by means of comparing the curves with the equation which is based on the reasonable theory.  We referred the mechanisms studied by R. Akolkar and U.Landau ²⁾ and modified their theory to apply to flat disk electrodes.
   We obtained the following equation(1) expressing the relationship between time (t) and potential(η) by deforming their equations.  The typical result of fitting this equation with the experimentally obtained curve is shown in Fig.5.  The calculation was made using the conventional tool (Solver in Excel).  The curve was fitted well.  This indicates that their mechanisms can be applied for monitoring additives.  We acquired the parameters, i_0,I, i_0,A and C^b_A / Γ_I from this study as follows:
      i_0,I : 0.037 mA/cm²  ,  i_0,A :  0.129 mA/cm² ,  C^b_A / Γ_I  : 1219/cm
For monitoring in the practical operation, the change of these values could be the indexes of the additive control.
   Additionally it is supposed that the cuprous ions produced during copper deposition on the electrode surface should work as another accelerator and affect the reaction depending on the current density, that is, the displacement rate constant k becomes the function of current density.  In that case, equation(1) should be modified.  We are investigating the applicability of the modification.

4. Summary
   The methods obtaining and analyzing time-potential curves are effective for practical additives control.  The curves express the behaviors of the additives affected by the various conditions, such as degraded by-products of additives.
   The features of the curves can be analyzed to characterize numerically.  Moreover, it was found that good curve fitting of the acquired curves with the equation based on the theory proposed by R. Akolkar et al. could be done and the meaningful parameters could be obtained.

References
1) T. Okubo, Y. Mizuno, K. Naoi, J. Japan. Ins. Electronic Packaging, Vol.8,No.4, p318-324(2005). (in Japanese)
2) R. Akolkar, U. Landau, J. Electrochem. Soc., 156(9), D351-D359 (2009).