Characterisation of the Oxidation Kinetics of Thin, Low Temperature, Electroless Plated Copper Films

  • PDF / 1,165,379 Bytes
  • 6 Pages / 414.72 x 648 pts Page_size
  • 1 Downloads / 199 Views

DOWNLOAD

REPORT


INTRODUCTION Copper interconnect will play an important role in future ultra large scale integrated (ULSI) devices because of its lower bulk resistivity and superior resistance to electromigration as

compared to aluminium and its alloys [1]. While several deposition techniques are under

development considerable research has now focused on electroless copper metallisation as it is a low temperature process capable of depositing copper on organic interlevel dielectrics [2]. The morphology, chemical state and oxidation behaviour of electroless copper films deposited from two different baths were characterised using SEM, AFM, XRD, SIMS and SE. In particular this work focuses on the characterisation of the oxidation of as-deposited electroless copper films as Cu is known to oxidise at a significant rate at temperatures as low as 100'C [1, 35]. The antioxidising effects of a selectively deposited organic film is also investigated. The first step in the oxidation of copper is the formation of cuprous oxide (Cu20) which then oxidises further forming cupric oxide (CuO). It is generally considered that the rate limiting step of this reaction is the diffusion of copper ions through the Cu20 scale to the scale/oxygen boundary and that the formation of CuO does not effect the overall reaction rate as it proceeds more quickly [1,6]. Although there is not strict agreement between much of the published data it is generally agreed that the low temperature oxidation of copper follows a quasi-logarithmic rate law [7]. EXPERIMENTAL Copper films were prepared from two different electroless baths, a commercial formaldehyde (HCHO)-based bath (Cupro-Thick 84, Alfachimici) and a formaldehyde-free (dimethylamine borane (DMAB)-based) bath developed at NMRC [8,9]. In both cases the copper was deposited onto TiN (0. 1Pm)/SiO 2 (0.1 ltm)/Si substrates. The HCHO-based bath operated at 35'C and the DMAB-based bath at 55'C with plating rates of 31rm min- and 20nm min-1 respectively. 353 Mat. Res. Soc. Symp. Proc. Vol. 406 ©1996 Materials Research Society

Half of the wafer on which the commercial copper was deposited was coated with a commercial oxidation inhibitor (Orgasol, Alfachimici). The coating could be removed by a weak acid solution (0.3mol dm-3 citric acid). The morphology of the as-deposited Cu films was determined using scanning electron microscopy (SEM) (Hitachi S-4000) and atomic force microscopy (Topometrix TMX 1010 Explorer). The oxidation products present on the copper films were determined using X-Ray diffraction (XRD) (Philips PW 3710 MPD). Secondary Ion mass spectrometry (SIMS) (Cameca, IMS-3F) was employed to determine copper film purity. The as-deposited copper film samples were heated in an oven at 5°C min- 1 in a nitrogen atmosphere until the pre-set isothermal temperature was reached. The oven was flushed with dry synthetic air for the duration of the isothermal oxidation as detailed in table 1. SE measurements were performed using a phase modulated variable angle spectroscopic ellipsometer having a measurement spot size