Development of novel alternative chemistry processes for dielectric etch applications

Thesis (Ph.D.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 2000.Includes bibliographical references.The removal of dielectric films in semiconductor processing relies almost exclusively on the use of perfluorocompounds (PFCs), which are suspected global warming agents. The two applications in semiconductor manufacture that account for the largest use and emissions of PFCs are the patterning of dielectric films and the cleaning of dielectric film plasma enhanced chemical vapor deposition (PECVD) chambers. The work discussed in the author's Ph.D. thesis was conducted as part of a project whose goal is to identify and develop novel replacement etchants for these applications. The focus of the author's Ph.D. thesis is the patterning application. The research discussed in this document constitutes a follow up to the author's M.S. thesis, which discussed the initial stages of this project. These stages consisted primarily of preliminary screening tests involving a class of chemistries which was expected to be promising from a process standpoint at an early point in the project. The work carried out subsequently covered a far greater scope of activities and included additional preliminary screening tests with chemistries that were not covered by the author's M.S. thesis as well as extensive concept-and-feasibility tests and subsequent process development efforts using several of the more promising chemistries in a dielectric wafer patterning application. Much of this experimental work had been carried out in collaboration with industrial partners belonging to the semiconductor manufacturer, equipment supplier, and gas supplier communities. These tests were carried out on process tools housed both within MTL's Integrated Circuits Laboratory and at an industrial location, namely Motorola Inc.'s Advanced Products Research and Development Laboratory (APRDL). The project to identify and develop alternative chemistries for dielectric film removal applications is continuing after the completion of the author's thesis, with subsequent studies that will build on the results of the work done to date. The research presented in this document involved the evaluation of fluorinated compounds belonging to three principal families of modified fluorocarbon molecules: hydrofluorocarbons (HFCs), iodofluorocarbons (IFCs), and unsaturated fluorocarbons (UFCs). In addition, other chemistries, namely trifluoroacetic anhydride (TFAA), oxalyl fluoride, and octafluorotetrahydrofuran, were also examined. The focus of much of the work was on the etching of patterned silicon oxide films in back-end-of-the-line (BEOL) applications such as via etch. In its mature phases, the work conducted relied on a two pronged approach toward evaluating new etchants: characterization of their process performance and characterization of their process emissions prior to release into the atmosphere. Cross-sectional scanning electron microscopy (SEM) was the principal means of process performance characterization, whereas Fourier transform infrared (FTIR) spectroscopy was the principal technique employed for effluent characterization. At appropriate times, other diagnostic techniques, namely x-ray photoelectron spectroscopy (XPS), time-of-flight secondary ion mass spectrometry (TOF-SIMS), Auger electron spectroscopy (AES), and quadrupole mass spectrometry (QMS) were also used for film or effluent characterization. Within the HFC family, candidates were identified that exhibited generally good process results with emissions reductions ranging from -40% into the 70% range relative to a PFC based reference. More comprehensive tests with IFC compounds demonstrated that emissions reductions in the 80% range are attainable for working processes. Good performance was obtained with respect to some, but not all, key process metrics, demonstrating the potential utility of IFCs in certain dielectric etch applications, but also indicating that there were significant limitations to their use, stemming mostly from selectivity problems. In tests with UFC compounds, emissions reductions on the order of 85%, combined with good process performance, were obtained. This family of compounds showed the greatest promise from both an emissions standpoint and a process standpoint. In particular, compounds in this family showed very good mask layer and stop layer selectivity, in addition to good etch rates and good profile control. It is particularly encouraging that the use of some of these compounds, in addition to offering emissions reductions, may, in fact, offer a process advantage over conventional chemistries in applications requiring high selectivity. At the time of this writing, unsaturated fluorocarbons are viewed as a major avenue for further exploration within the ongoing PFC alternatives project.by Simon Martin Karecki.Ph.D