Solid State Phenomena Vols. 145-146

Abstract: The integration of CoWP and CoWB self-aligned barriers (SAB) for 32 nm technology nodes allows improving copper interconnections reliability [1, 3]. However the introduction of such materials in copper interconnection levels drives new challenges for plasma dry etch and wet clean processes. Indeed, during the post-via-etch cleaning step, cobalt and copper can be altered by corrosion. Moreover, a galvanic coupling between cobalt, the major component of SAB, and copper can thermodynamically occur. In this way, the cleaning solution acts as ionic medium providing a contact between the two metals. Thus, both metals polarize to a mixed potential comprised between the individual open circuit potentials (OCP) of cobalt and copper. As a result, the less noble metal can suffer from accelerated corrosion, and the more noble metal corrodes with slower rate. According to thermodynamic aspects, cobalt in contact with copper is the less noble metal. Consequently, Co is susceptible to undergo galvanic corrosion which may enhance the dissolution of the SAB.

Abstract: The subject of this report is the characterization of plasma etch residues after a metal etch process with Cl2/BCl3 etch gases. One of the interactive factors in the removability of the residues is the photo-mask removal process (DSQ). Depending on the DSQ process the molecular structure of the residues will differ. For our findings, we used laser spectroscopy and Fourier-transformed infrared spectroscopy to obtain information about the degree of the cross-linking of the molecular structure of residues in a post-metal etch cleaning process. The post-etch cleaning is important for removing residues remaining after the metal structuring process. The main goal is to use emission spectroscopy for studying the compounds of the dry-etch related residues. Finally, it was shown that small variations in wafer treatment directly after dry-etching results in different solubilities of residues in HDA (hydroxylamine) based solutions. [1]

Abstract: Driven by cost saving pressure and new regulations such as REACH, the imple¬mentation of cost efficient, EHS-friendly alternatives is in full swing. There is a clear trend in the plasma etch residue (PER) cleaning field to replace the currently used, expensive and harmful organic PER-removers by EHS-friendly aqueous alternatives. Striking advantages of fluorine free aqueous removal are reported1. The goals of present work -which was carried out in cooperation of BASF and Infineon-, was to examine the dependence of residue constitution on etch and strip conditions and to find a way to adapt also most critical cleaning challenges for EHS-friendly PER remover usage. For that reason a selected aluminum interconnect metal layer with difficult removable residues was chosen. XPS is well proved as an appropriate method for selective sidewall and top/bottom residue analysis2. Our analytical work was focused on defining fingerprints for the etch residues, concerning their elemental distribution, and tracking their changes during the different process steps and conditions.

Abstract: The borderless (BDL) vias landing on the metal lines were demanded in high-density flash memory devices due to the reduced die size, compared with non-borderless (NBDL) vias. IMD material recess in un-landed via dry etching and the Al-Cu metal line undercut caused by post-etch cleaning may lead to via resistance Rv high and EM reliability issue. In this work, the Al-Cu loss mechanism in BDL-via cleaning by using amine- and fluoride-based chemicals were studied in 58nm flash devices. We found that the Al-Cu undercut can be minimized in HDA containing solvent by controlling the formation of Al2O3/Al/HDA gel-like material to the less amounts; the shorter HDA processing time with sufficient IPA rinse generated the thinner gel-like layer leading to the less Al-Cu undercut in DIW rinse step. For TMAH/H2O2-containing semi-aqueous solution, the metal contained etched-residues can be removed by reaction of OH- with Al to form dissolved Al(OH)4-; addition of H2O2 will form a oxidized passivation layer on the metal surface to prevent from metal corrosion. On the other hand, the Al/Al2O3 contained polymer was removed by fluoride-based solvent or DSP+ through the reaction with [F-] to form the DIW-dissolved AlFx by-products; while the dilute H2SO4 and H2O2 in DSP+ will oxidize Al to form Al2O3 that will be further etched by HF, the severe Al-Cu undercut was found in DSP+.

Abstract: Brush scrubbing has been widely used in post chemical mechanical planarization (CMP) applications to remove contaminations, such as slurry residues and particles, from the wafer surface. During brush scrubbing, particle removal results from direct contact between a soft poly vinyl alcohol (PVA) brush and the wafer surface in which the brush asperities engulf the particles while the rotating motion of the brush, as well as the cleaning fluid at the surface, dislodge and carry the particles away from the wafer. The cleaning performance of brush scrubbing depends heavily on the choice of the cleaning solution and brush scrubber kinematics. In this work, the effect of various cleaning solutions and brush scrubber kinematics on the frictional attributes of post copper CMP cleaning process was investigated.

Abstract: Cu (copper) has been widely used for interconnection structure in integrated circuits because of its properties such as a low resistivity and high resistance to electromigration when compared with aluminum [1, 2]. Damascene process for the interconnection structure utilizes 2-steps CMP (chemical mechanical polishing). After 2-steps CMP process, many abrasive particles leave on the wafer surface, which should be removed in post-Cu CMP cleaning process. Cleaning efficiency affects directly on the subsequent process and device yield [3]. Therefore, cleaning of abrasive particles is the critical issue in semiconductor manufacturing.

Abstract: In damascene architecture, widely used both in flash memories and in DRAM as interconnect scheme since 90 nm node, copper surface is exposed after via etch. A deep understanding of the effect of different wet cleanings on Cu surface is therefore mandatory, not only to ensure an efficient post etch polymer removal, but also to provide a better surface termination, capable to minimize Cu oxidation kinetic and to reduce the growth of Cu-rich precipitates which may negatively effect contact resistance. In this work we have analyzed the Cu surface after processing with several cleaning chemistries -often present in BEOL cleaning processes- using XPS (X-ray Photoelectron Spectroscopy) and ToF-SIMS (Time of Flight – Secondary Ion Mass Spectroscopy), fast and powerful techniques widely used in Cu surface characterization [1]. In addition, the evolution of the surface with storage time has been monitored using the same techniques, in order to better understand the effect of the different cleaning chemistries. XPS has been proven to be very sensitive to monitor Cu oxidation, while ToF-SIMS has been used to reveal organic species adsorbed on the surface.

Abstract: The development of robust integration processes for low-dielectric-constant materials is critical in order to meet the ITRS timeline. For 45 nm and beyond, the roadmap dictates use of an advanced dielectric material with a bulk dielectric constant below 2.5. These materials are produced through the introduction of nanometer-scale porosity into an OSG skeleton. The pore size, morphology, and interconnectivity is controlled by the choice of OSG precursor, pore former, and process conditions. During integration, liquids can be absorbed into the pore network (e.g. during the polishing and cleaning process steps, resulting in a degradation of the electrical and mechanical properties [1-3]. The objective of this paper is to evaluate the impact of the CMP and post-CMP cleaning process steps on cohesive fracture in a porous OSG dielectric and the adhesive fracture of SiCN cap with porous OSG and copper films.

Abstract: As technology node progressing, ultra low-k film has been implemented to reduce RC delay in LSI circuit. A fluorocarbon (CFx) film is proposed as foreground ultra low-k film because of non-porous structure [1]. Although CFx film is expected to be stable for its structure advantage, damage-less process is anticipated to avoid dielectric constants change in subsequence process steps [2]. CMP and post CMP process are concerned to bring damage on devices, so the effect of post CMP cleaning solutions on CFx structure and electric property is evaluated.

Abstract: At critical dimensions of 65nm and lower, the tolerances for yield impacting “killer” particle defects become ever tighter and this is especially true for Cu PCMP cleaning. This is a critical step in reducing wafer defectivity and metal contamination directly impacting yield and reliability. A successful candidate must exhibit a number of additional attributes not least of which are compatibility to Cu while selectively removing copper oxides (CuOx), and be fully compatible with porous ULK and barrier materials. Conventionally PCMP cleans typically require two different acidic or alkaline solutions containing a surfactant to drive particle removal and surface metal preparation; however this is not without its limitations e.g. crack propagation, water absorption, low-k wetting, water marks and galvanic corrosion. It would be advantageous if Cu PCMP could be performed with a single neutral pH cleaning solution. The paper will present results of work conducted on a neutral, single step Cu Post CMP (PCMP) cleaning solution, EKC PCMP5510™, and will discuss the influences of different process parameters in optimising defect removal from Cu and Ultra Low-K (ULK) dielectric surfaces (K=2.5) thus highlighting the fundamental differences in cleaning either surface.