Issue link: https://iconnect007.uberflip.com/i/712650
August 2016 • The PCB Magazine 89 FACTORS AFFECTING THE ADHESION OF THIN FILM COPPER ON POLYIMIDE 1−2 microns thick would be best for masking off the sintered layer, electroplating a pattern, removing the mask, and etching off the un- wanted copper. Since etching is isotropic, the sides of the plated traces will etch as well at the tops, and thinner traces give less under-cutting and straighter side walls. Using the ultra-thin film process, straight side walls of features on the etched film can be achieved, as shown in Figure 4. Present methods of creating these 1−2 micron films involve sputtering or etching thick foil laminated layers and are very expensive. In the present paper, we discuss the adhe- sion of the thin copper films on flexible poly- imide substrates created by ultra-thin film pro- cess, and the possible factors to improve the adhesion. Methodology The adhesion of ultra-thin films was tested by the ASTM 3359 cross-hatched tape test in which a pattern of horizontal and vertical cuts spaced 1 mm apart are made in the copper layer by a blade (a razor blade was used). A special tape is adhered to the cross-hatched area and peeled off at about 180°. To test the adhesion of the patterned, plat- ed and etched copper layers, a peel test follow- ing IPC-TM-650-2.4.8 is routinely used in the printed circuit industry. Here special coupons >30 microns tall are created. They are 1 mm wide by up to 96 mm long, with a wider tab at the end for attaching by a clamp to a force gauge. The trace is peeled in a controlled man- ner at 90 degrees to the surface of the poly- imide and the force measured as a function of peel length. A photo of such a sample is shown in Figure 5. Data and Results One of the attractive features of this ultra- thin film process has been that the coated, dried and sintered 0.5-micron copper layer adheres very well to polyimide. Due to the fine thick- ness of the films, the ASTM 3359 cross-hatched tape test has been applied to test the adhesion of the coatings. In the present study, the tested thin film samples were prepared by a lab scale preparation, i.e., Mayer bar coating, vacuum dry and single flash lamp sintering, and it was found that none of the sintered copper is re- moved by the tape. However, for the patterned, plated and etched copper layers, it was found that these 1 mm wide traces have highly variable adhesion. A peel force of about 10 N/cm is acceptable in the industry, and there have been samples that ranged from 0 N/cm (came off in the etching bath) to 8 N/m. So, a sintered copper layer was used that adheres very well to polyimide, but when that layer is masked, plated, and etched, the resulting (>30 micron) layer sometimes does not adhere very well. It is instructive to point out that the plated etched area is >30 mi- crons thick because the peel coupons must be that thick to avoid tearing when connected to the peel test apparatus and pulled. In order to understand the possible sources for this variable adhesion that are seen for plat- ed and etched samples, we have investigated a number of factors which might affect the ad- hesion in our system. First, we investigated the flash sintering conditions. Well-sintered layers usually conduct the best and have the best ad- hesion. Additionally, we have mentioned al- ready that sintering large areas of coated ink by overlapping repetitive flashes from a single bulb shows a darker band in the overlap area, and it has been found that the overlap area does not exhibit as good adhesion as the more copper col- ored area from a single flash (data not shown). The SEM images from the copper colored area and darker band (Figure 6) clearly showed the Figure 5: A plated, etched sample of nano copper ink on polyimide showing the peel strength coupon.