通过湿化学清洗结合物理力预处理去除离子植入后光刻胶

A combination of wet chemistry and high-velocity solid CO2 aerosol  pre-treatment was used to remove ion-implanted resist from  patterned Si structures. The top resist surface is modified by heavy  ion-implantation (1x1016 As atoms/cm2 , 40keV) forming a crust  ~100nm thick. The aerosol treatment, which is aimed at breaking  and partially removing the crusted top layer of implanted resist,  produces circular openings in this surface, with diameters on the  order of tens to hundreds of micrometers. No resist is detected by  either optical microscopy or localized SEM inspection, after  immersion in hot H2SO4/H2O2 mixture (SPM), on aerosol-pretreated wafer fragments. In contrast, resist residue is still observed  on similarly pre-treated specimens, after immersion in either Nmethyl pyrrolidone or Microstrip2001™, even under ultrasonic  agitation. Overall, the aerosol pre-treatment is a key step in reducing  the amount of resist residue, for both aqueous-based and solventbased wet cleaning. Full-wafer defect-count results, post aerosol  treatment, are reported for poly-Si line widths down to 40 nm.

Introduction

One of the most critical challenges for future generation devices is the effective  removal of post-ion-implant photoresist, without damage to underlying gate stack  structures (1). Additional requirements for a viable resist strip process must include  complete removal of any resist residues, no silicon substrate or dopant loss, and  compatibility with novel metal-gate and high-permittivity (high-k) materials. Resist is  typically employed as block mask during ion-implantation for front-end-of-line (FEOL)  CMOS device processing flows. Since elevated ion doses and energies are necessary for  increased device performance, the top resist surface undergoes chemical modification  under the ion flux, resulting in the formation of a “crust” which is very difficult to  remove at the required high selectivity specifications, by either wet or dry methods. For  example, photoresist stripping with aggressive, fluorine-based plasma chemistries leads  to unacceptably high silicon substrate and dopant loss. At the same time, wet-chemistry  treatments alone can not effectively attack or penetrate the top crust, thus being  unsuccessful in removing heavily ion-implanted resist layers. Alternative resist strip  methods have been proposed, such as a combination of physical force pre-treatments  followed by more traditional wet cleaning steps. For example, the removal of blanket,ion-implanted deep-ultraviolet (DUV) resist (As, 1x1016 atoms/cm2 , 80keV) from bare Si  wafers, using a combination of cryogenic CO2 and wet cleaning techniques has been  reported (2). In addition, other previous studies have also reported removal of ionimplanted resist at various doses and energies, using different methodologies such as  supercritical CO2 combined with chemical additive formulations (3), a piranha-based  process reaching 200°C on-wafer (4), a single-wafer process using sulfuric-acid-based  mixtures (5) and an all-dry process consisting of a plasma strip step followed by a  pulsating CO2 cryogenic aerosol treatment (6).  This paper reports on ion-implanted resist removal studies using a combination of  high-velocity solid CO2 aerosol pre-treatment and wet chemistry. Resist specimens are  characterized post wet and aerosol treatments using Optical Microscopy, SEM, HighResolution Profilometry (HRP) and X-ray Photoelectron Spectroscopy (XPS). The  effectiveness of SPM vs. solvent-based chemistries for implanted resist removal with and  without physical force is also evaluated.

Experimental

Both 200mm and 300mm wafers were used for these experiments. The 200mm Si  wafers were coated with KrF DUV resist (JSR) in a checkerboard pattern. This allowed  for the presence on the same wafer of three distinct surface regions: (i) blanket resist on  top of patterned poly-Si, (ii) patterned resist on top of patterned poly-Si and (iii) exposed  patterned poly-Si (i.e. no resist present). All resist-coated wafers were subsequently ionimplanted with a dose of 1x1016 As atoms/cm2 , at 40keV. The post-implant resist  thickness was ~350nm. The most aggressively scaled poly-Si lines on the 200 mm  wafers were nominally 70 nm wide and 100 nm tall. The 300mm patterned Si wafers,  which were used for defect-count studies in the absence of resist, had poly-Si lines with a  wide range of width and pitch values, down to 40nm wide and 100nm tall. Wafers were  prepared at IMEC and were subsequently sent to the BOC Eco-Snow location, where the  high-velocity solid aerosol pre-treatments were performed. The wafers were then sent  back to IMEC, where all characterization and wet strip experiments post-aerosoltreatment were performed on wafer fragments, with the exception of full-wafer defectcount inspection, which was done on a 300mm KLA-Tencor 2351 tool.

Figure 1: Raman spectra obtained for DUV  (atoms/cm2) resist specimens after ion-implantation with As a 1x1016 doses from 1x1014 to 1x1016 atoms/cm2 at 40  b 1x1015 keV.

Figure 2: HighResist Resolution crust Profilometry (HRP)  3D scan of an  implanted resist layer  opening formed by  solid aerosol treatment.  The insert shows a topdown optical image of  Underlying resist the same area.