We evaluated the removal of polymers with various chemical structures using wet ozone, and investigated the reaction mechanism between wet ozone and polymers using fourier-transform infrared (FT-IR) and in situ FT-IR. The removal rate of poly(vinyl phenol) (PVP), which has a carbon–carbon double bond (C=C) in the side chain was lower than that of the novolak resin, which has C=C in the main chain. Poly(methyl methacrylate) (PMMA), which has no C=C, was not removed. It was considered that the ozone reaction is an electrophilic reaction, and the wet ozone should react with C=C with ease. The removal rate of PVP with rinsing was higher than that without rinsing. This result indicates that the reaction products remain on the Si wafer. However, in the novolak resin, there was no difference between with and without rinsing. It was considered that the main chain of the novolak resin was decomposed to gas by the reaction with wet ozone. In the FT-IR measurement of PVP, the peak intensity of C=O stretching of carboxylic acid increased with increasing wet ozone processing time. However, in the novolak resin, there was no difference between with and without rinsing. Moreover, the peak intensity of the C=O stretching of carboxylic acid did not increase with increasing wet ozone processing time after 10 s of wet ozone processing time. From the result of in situ FT-IR, in the removal of the novolak resin using wet ozone, the main chain of the novolak resin was decomposed, and the reaction products of the wet ozone and novolak resin (low-molecular-weight carboxylic acid) should change to CO2.

We have investigated the removal of novolac resists into which B and P ions had been implanted with a dose of 5×1014 atoms/cm2 at acceleration energies of 10, 70, and 150 keV (ion-implanted resists), using wet ozone. Also, we confirmed the presence of the surface hardened layer of ion-implanted resists by cross-sectional observation using scanning electron microscopy (SEM), the stripping of the hardened layer using chemicals, and the measurement of the plastic-deformation hardness of the resists by nanoindentation. The removal rate for a resist using wet ozone decreased with increasing acceleration energy because the hardness of the resist increased with increasing acceleration energy. Moreover, we clarified by time-of-flight secondary ion mass spectrometry (TOF-SIMS), that the ion intensity of C10H- (m/z 121.01) for the hydrocarbon component, which has the unsaturated bonds as a component of the surface hardened layer increased with increasing acceleration energy. Cresol novolac resin was destroyed and carbonized by ion implantation. Therefore, the removal of ion-implanted resists became difficult with increasing acceleration energy.

We have investigated the removal of poly vinyl phenol (PVP) into which was B, P and As ions implanted with a dose of 5×10 13~5 ×10 15 atoms/cm 2 at 70keV, using wet ozone. Also, we investigated the thickness of altered layer of ion-implanted PVP using Secondary Ion Mass Spectrometry (SIMS). We investigated the chemical reactivity of wet ozone and ion-implanted PVP by FT-IR. The removability of ion-implanted PVP using wet ozone decreased with increasing dose because the degrees of alteration of benzene ring and O-H group increase with increasing dose. From the results of SIMS, the thickness of altered layer of B-ion-implanted PVP was 387nm, that of P-ion-implanted PVP was 232nm, and that of As-ion-implanted PVP was 142nm. However, the degree of alteration should increase in order of B, P and As, because any samples were implanted at same acceleration energy (70keV). FT-IR indicated that benzene ring content percentage was almost same at removed and not removed. On the other hand, in O-H group, there is a clear difference at removed and not removed, O-H group content percentage decreased drastically at implanted PVP with a dose of 5×10 14 atoms/cm 2 over. Therefore, the removability of ion-implanted PVP using wet ozone depends on O-H group content than that of benzene ring.

No abstract

No abstract