Electroless Deposition of Cu Thin Films with CuCl[sub 2]-HNO[sub 3] Based Chemistry: II. Kinetics and Microstructure

Abstract

This paper describes the kinetics involved in the acid-based electroless Cu deposition system employing CuCl 2 -HNO 3 chemistry in a HF-NH 4 F buffer solution. The rate equation is set up as a function of concentrations of active chemical components involved, and rate orders are determined to evaluate the contribution from each component. The deposition rate of Cu is found most sensitive to variation in ͓Cl Ϫ ͔ concentrations, followed by that of ͓Cu 2ϩ ͔ and ͓F Ϫ ͔. The activation energy derived from deposition at different temperatures is 0.445 eV. Grain size of deposited Cu films is influenced strongly by deposition rate. Electrical resistivity of Cu films is dominated by the amount of point defects and microvoids present right after deposition, and by grain size after a 300°C anneal.Electroless plating provides a viable alternative to electroplating for Cu thin-film deposition in the era of Cu-based multilevel interconnect for microelectronic devices. 1 Besides its capability of filling high-aspect-ratio pattern structure, electroless plating offers additional advantages such as low cost and simpler chemistry.Most existing electroless Cu plating processes adopt base solutions with copper sulfate as the Cu ion source, ethylenedinitroloacetate ͑EDTA͒ as the complexing agent, formaldehyde as reducing agents, tetramethylammonium hydroxide ͑TMAH͒ or alkali solution as pH buffer, and some surfactant and stabilizer added to improve step coverage and bath stability, respectively. This chemical formulation has been studied extensively over the past twenty years and has been applied to the manufacture of circuit boards.In the first part of this study, 2 we proposed an acid-based CuCl 2 -HNO 3 electroless Cu plating solution with HF-NH 4 F as pH buffer. An optimum deposition rate of ϳ2700 Å/min directly on TaN substrate and a resistivity of 2.35 ⍀ cm can be achieved at pH 4.70. CuCl 2 , when dissolved in the solution, provides the ͓Cu 2ϩ ͔ ions, which are catalyzed by HNO 3 on TaN surface and reduced by Si to form Cu deposits. ͓Cl Ϫ ͔ and ͓F Ϫ ͔ ions act as complexing agents and transmit the electrons from Si to ͓Cu 2ϩ ͔. Cu deposition rate exhibits a decreasing trend with increasing ͓HNO 3 ͔ concentration in the bath. Adding ͓HF͔ or ͓NH 4 F͔ into the bath, on the other hand, enhances the deposition rate to a peak level, after which it saturates and declines.As an extension of our previous study, we explore the kinetics and material issues of the system in this contribution. Concentration of the active components and temperature used in the system are varied in order to deduce the reaction rate orders and activation energy. Evolution of grain size and resistivity with ͓Cu 2ϩ ͔ concentration in the chemical bath are analyzed. The effects of postdeposition anneal on resistivity, microstructure, and preferred crystal orientation of Cu thin films are also investigated. ExperimentalDetermination of reaction rates.-Chemical components and substrate materials used for the Cu deposition experiment basically follow those from ou...

Behavior of Electroless Cu Deposition in CuSO[sub 4]-HF Solution

“…It is noticed that the activation energy at d ϭ 0.5 mm ͑0.486 eV͒ is close to that of reported acid-based electroless Cu deposition ͑0.445 eV͒. 13 Hence it is deduced that the process of electron diffusion in solution is dominant in the electron transmission for its acid-based electroless Cu deposition. 12 Dependence of c 2 on plating time and bath temperature.-In Eq.…”

“…[12][13][14] W. T. Tseng et al 12,13 produced Cu film on TaN barrier successfully with CuCl 2 -HNO 3 based chemistry. The electroless Cu deposition was performed in a HF-NH 4 F buffer solution and ͓F Ϫ ͔ and ͓NO 3 Ϫ ͔ ions dissolved Si on back and bevel of the wafer.…”

Effects of Pd activation on the self annealing of electroless copper deposition using Co(II)–ethylenediamine as a reducing agent

“…This occurred because the increase in the Pd ion concentration accelerated the Pd nucleation rate on TiN surface by lowering the critical radius of nuclei r * . 16 The highest Pd particle population with small sizes was observed at 0.10 g / L. The size increased exponentially with the increase in the Pd activation time. Decrease in the number of Pd particles per unit area, as a function of time, in spite of the increase in the particle size is due to agglomeration of particles.…”

Electroless Ni–Mo–P diffusion barriers with Pd-activated self-assembled monolayer on SiO2

“…The scotch tape test is usually employed to evaluate adhesion of the films on the substrates [31], and it was repeated three times on each specimen. Those without any visible Ni-Mo-P damage after tests are classified as excellent, and those with some spots left on the film after tests are defined as good, and those which peeled off easily are decided as poor [32][33]. In this paper, the adhesion tests were performed by the Cu/Ni-Mo-P/SiO 2 /Si and Ni-Mo-P/SiO 2 /Si specimens.…”