Time of flight secondary ion mass spectroscopy investigation of ultralow-k dielectric modifications in hydrogen and deuterium plasmas

Abstract

Photoresist stripping processes based on hydrogen reductive chemistry have emerged as attractive replacements of the conventional treatments performed using O 2 plasma. H 2 discharges and plasma processes based on gas mixtures containing H 2 provide, in fact, adequate stripping rates while reducing the chemical modifications of the porous ultralow-k ͑ULK͒ dielectric materials employed in interconnect technology. In this work the authors investigate the chemical modification of a number of organosilicate low-k dielectrics interacting with remote H 2 plasma. The materials include both spin-on and plasma-enhanced chemical vapor deposited dielectrics with k, porosity, and chemical composition spanning on a broad range of values. The mechanisms of plasma/dielectric reaction and the depth of interaction are evaluated by means of time of flight secondary ion mass spectrometry. In order to study the nature of the plasma/dielectric interactions and the chemical modifications introduced in the ULK materials, deuterium ͑D 2 ͒ is used to substitute for H 2 in the discharge. Although the plasma-generated radicals interact strongly with the materials, negligible stoichiometry modifications are observed during exposure of silica-rich ULK materials. On the other hand, the dielectrics with higher carbon content undergo major compositional changes. The chemical damage introduced by H 2 -based discharges depends strongly on the formulation of the organosilicate material. The substrate temperature is also seen to affect the influx of plasma species. This acts on the extent of material modifications. Nevertheless, the nature of plasma/dielectric interactions does not vary for processes performed in the temperature range of 200-300°C.

Mechanisms of porous dielectric film modification induced by reducing and oxidizing ash plasmas

“…Worsley et al suggest that, at elevated temperatures, the concentration of H species chemisorbed at the surface decreases due a lower sticking coefficient and that the material modification is strongly dependent on the porosity rate and the interconnectivity of the pores. 21 As for Lazzeri et al, 28 the chemical damage introduced by H 2 -based discharges depends strongly on the formulation of the organosilicate material. The substrate temperature is also seen to affect the influx of plasma species ͑extent of material modifications͒.…”

“…wSi -O -Si w + 2H → w Si -H + Si -OH ͑⌬H r = + 42 kj mol −1 at 298 K͒, ͑4͒ 21,25,26,28 All these studies point out that the film modification is temperature dependent and is also correlated to the density of the H reactive species concentration in the plasma. 21 The film damage is minimized by increasing the substrate temperature, 25,26 while the damage is amplified with the reactive species concentration.…”

Toward Successful Integration of Porous Low-k Materials: Strategies Addressing Plasma Damage

“…In the case of ashing chemistries, static TOF-SIMS was used to study the nature of the top surface, while dynamic TOF-SIMS provided additional information concerning the lack or presence of additional damage in the bulk of the dielectric material. 54,55 Similarly, the mechanism of fluorocarbon-based etch plasmas could be established using static SIMS, 56,57 while confirming bulk dielectric damage in the form of carbon depletion by dynamic SIMS. Minimum plasma damage and pore sealing by oxygen-free fluorocarbon plasmas could also by corroborated using TOF-SIMS among other characterization techniques.…”

Improved characterization of Fourier transform infrared spectra analysis for post-etched ultra-low-κ SiOCH dielectric using chemometric methods

“…Recently, the PCA was applied to study changes of ULK films after plasma in combination with ToFSIMS. 14 It is shown that the amount of carbon loss due to plasma treatments is visualized by typical FTIR peaks ͑Si-CH 3 , Si-C, C-Si-O, and Si-O-Si cage/network͒. Here, PCA was used to study structural bond change, since this method is sensitive variations of FTIR peaks.…”