Issue link: https://iconnect007.uberflip.com/i/688506
June 2016 • The PCB Magazine 35 HIGH-THROW ELECTROLESS COPPER—NEW OPPORTUNITIES FOR IC SUBSTRATES AND HDI MANUFACTURING According to the required measurement points, the throwing power is calculated as follows: Equation 2: Throwing power measurement at different areas in the BMV. In order to achieve a sufficient measure- ment resolution, it is mandatory to apply SEM for evaluation of the cross sections. The SEM settings used for the thickness measurement are standardized to increase the comparabil- ity between different locations and SEM types. Nevertheless, the measurement is still operator dependent but according to intensive testing of the method and statistical evaluations, the ex- pected error is in an acceptable range. The appli- cation of focused ion beam (FIB) cutting would be a potential alternative to SEM, but due to the immense cost and throughput constraints for the measurement it is not a viable option. The presented method A is suited to measure the throwing power directly on the base mate- rial but not on copper because etching or elec- tro polishing is required in order to distinguish between different copper layers (e.g., electroless copper vs. the capture pad) and the method does not allow these preparation techniques. As described, etching or electro polishing attacks the electroless copper layer quite strongly, that is why two copper protection layers are electro- lytically plated in method B before and after the electroless copper deposit to be measured (Fig- ure 5). This build-up enables a stable electroless copper thickness even after etching and electro polishing and therefore provides an accurate method of electroless copper thickness mea- surement on the capture pad or on the copper clad surface. The measurement of the throwing power is performed analogue to method A by cross sec- tions and SEM images. Throwing Power Performance As discussed, throwing power is one of the most important performance criteria for next- generation electroless copper baths for both, IC substrates as well as HDI panels. On the one hand, a minimized electroless copper layer thick- ness on the surface of the laminate is required in order to reduce the differential etch and in- crease the achievable resolution. The electroless copper thickness in the wedge of the BMV on the other hand is limited to a certain thickness because of conductivity requirements of the subsequent via filling process step. A solution for these contradicting requirements is a high throwing power electroless copper process. Cur- rently, throwing power values of approx. 30% in the wedge and 70% on the capture pad of the BMV represent the typical performance in the industry for vertical IC substrate manufacturing whereas future design rules will require mini- mum 70% throwing power in the wedge and 100% on the capture pad. For horizontally pro- duced HDI boards in panel plating, the throw- ing power on the capture pad is approximately 30% and the requirements will increase due to the postulated technology shift to AMSAP. The throwing power performance is not only a function of the electroless copper bath, but also of the solution exchange influenced by the type of plating equipment, the aspect ratio (AR) of the BMV and the wedge formation respec- tively the AR of the wedge. Nevertheless, the chemical formulation and the process param- eters of an electroless copper bath significantly influence the throwing power performance. Figure 6 shows a typical BMV (60 µm width × 40 µm depth) on ABF GX-92 processed with Figure 5: Throwing power measurement (method B).