The surface of positive photoresist is hardened by the ion implantation process and becomes very difficult to remove thereafter. To investigate this phenomenon, the following aspects of photoresist were investigated: (i) the effectiveness of low-energy plasma ashing, (if) surface states by XPS, and (iii) outgassing during the ion implantation process. For photoresists without ion implantation, the ion bombardment energy dependence of the ashing rate was found to have two regions: the radical-mode ashing region and the reactive ion etching (RIE) mode ashing region. On the other hand, photoresist following an ion implantation process was found to be entirely covered with a hardened surface layer exhibiting only RIE-mode characteristics. This is why a plasma ashing process with low ion bombardment energy for suppression of damage on the wafer surface cannot work effectively in this case. When the photoresist is baked at a high temperature in a nitrogen ambience, the ratio between the carbon (C1~) peak area and the oxygen (O1,) peak area (C/O ratio) in the x-ray photoelectron spectroscopy data increases, indicating that carbonization has occurred. The ericial C/O ratio of carbonized photoresist that can be removed by sulfuric acid/hydrogen peroxide mix (SPM) is i0. However, an ion-implanted photoresist, with a C/O ratio of three, cannot be removed by SPM. Outgassing from the photoresist during the ion implantation process contains hydrogen as its main component. Our results indicate that an ion-implanted photoresist is difficult to remove because of an inactive high-polymer layer formed on the surface which has an extremely low hydrogen concentration, not because of carbonization in the surface. Therefore, the hydrogen concentration in the photoresist surface is considered to be critical to the efficiency of photoresist ashing. Since outgassing from photoresist during the ion implantation process contains hydrogen as its main component, the hardened surface layer must be supplied with hydrogen compounds in order to effectively remove the ion-implanted photoresist without damaging the substrate.

Formation of high-precision fine resist patterns has been achieved by an effective removal of dissolved resist polymers, that is, the reaction products of the development process. Also, it has been found that the reaction products give rise to the degradation of the resist contrast, resist sensitivity, and process margins. In order to remove the reaction products, two techniques have been employed, namely physical method of development with agitation (ultrasonic waves or with a stirrer) and chemical method of adding surfactant to the developer. By combining physical and chemical methods, the dissolved polymers have been effectively removed, thus allowing the resist to reveal its inherent patterning pertormance.

This is our reply to comments on our paper "Dose Pertubation by Wafer Charging During Ion Implantation." To our understanding, four issues have been raised: 1) There is nothing in the paper about the use of the electron shower. 2) Consistency between our results and those in [l].3) The contention that dose nonuniformity is an issue which has already been solved by the use of the electron shower, and that the current problem is the charge-up on a device structure level; and 4) the suitability of our implantation condition for CMOS fabrication.As for point l ) , we mention the use of the electron shower at the end of Section I11 in our paper.

The addition of small amounts of surfactant and hydrogen peroxide (H202) to the developer is shown to improve the performance of the photoresist development process. Exposed photoresist areas are dissolved more uniformly, the smoothness of Si surface is maintained, and carbon contamination during the development process is prevented. Ozone (03) treated ultrapure water rinsing at room temperature is an efficient way to remove the surfactant adsorbed on Si surface. A surfactant-added developer improves wettability on the photoresist surface, which leads to more uniform developing. The dissolution rate in the exposed photoresist is promoted and the etching rate of Si substrate is suppressed due to the effect of the additional surfactant. The only major disadvantage, due to the reaction of the surfactant in the strong alkali solution, is that the surfactant remains adsorbed on Si surface. H202 additions to the developer or ozone treated rinse water can remove the surfactant residues.

No abstract