Semiconductor devices have been achieved high performance by improving processing speed and integration degree due to downsize and multilayer wiring structure. Along with this tendency, the importance of CMP processing is increasing. In addition, due to many particles are generated after CMP process, post-CMP cleaning is also very important process. In the post-CMP cleaning, PVA (PolyVinyl Acetal) brush have been widely used. However, in this scrub cleaning, problems have been reported that many particles are generated from the PVA brush itself by adhesion of PVA brush onto surface and contaminated the wafer surface. Therefore, it is necessary to control adhesion between PVA brushes and semiconductor surfaces. In our early study, adhesion of PVA brush was evaluated by rotating torques or frictional forces. In this study, we tried to measure directly adhesion forces between PVA brush and several solid surfaces with a high response load cell. We believe that the measurement of the adhesion force can discriminate the material that is likely to adhere to the PVA brush. We compared the adhesion forces for two types of brushes with or without a skin layer. Moreover, we investigated whether adhesion force could be controlled by changing brush temperature. Because of the PVA brushes are porous polymer, mechanical properties change with its temperature. We consider that the temperature also affects adhesion force onto solid surfaces.

　In this study, we measured the adhesion forces between PVA brushes and solid surfaces in the atmosphere. The adhesion forces F_a were measured with a load cell. PVA brush was pressed against the test surfaces and releasing it at high speed after relaxing for a certain period of time t_w. We used three types of test surfaces, Glass, Polymethyl methacrylate (PMMA) and Polytetrafluoroethylene (PTFE). The contact angles were 28°, 61° and 97°, respectively. The temperature of test surfaces were controlled with a Peltier device and were set to 20 degrees to 50 degrees.

　Attached figure shows an example of relationship between the adhesion force F_a of PVA brush with skin layer and relaxation time t_w on the different surfaces. As shown in the figure, F_a increased with increase t_w in the all surface cases. We consider that this cause is real contact area increased as the t_w increased. In the dependence of test surfaces material, Glass and PTFE case showed similar adhesion force values. On the other hand, the forces of PMMA case showed larger values than other test surfaces. This result indicates that there are materials that strongly interacts with skin layer PVA brushes. In the experiments using without skin layer case, however, we could not measure the adhesion force in the all experimental conditions. This result suggests that without skin layer brush is maybe not contact onto surface and the adhesion force is strongly depended on the existence of skin layer. Next, we focus on temperature dependence of the adhesion forces. In the all test surface cases, adhesion force decreased with increase temperature. Particularly the adhesion force of the PMMA case was greatly decreased at 50 degrees. These results suggest that the temperature may control the adhesion force of PVA brush onto solid surface.