No abstract

No abstract

Both stress and atomic force microscopy (AFM) measurements were carried out in situ during potentiostatic electrodeposition of copper on gold in 0.05moldm−3 CuSO4 in 0.1moldm−3 H2SO4 with and without additives. With no additives, compressive stress generally developed initially and films subsequently underwent a compressive-to-tensile (C-T) transition. With increasing negative potential, the time for the C-T transition decreased rapidly as the rate of coalescence of nuclei (measured by AFM) increased rapidly. This is consistent with models that attribute the C-T transition to increasing tensile stress due to coalescence of nuclei. Furthermore, at a potential of −75mV (Cu∕Cu2+), where AFM showed very little coalescence of nuclei, no C-T transition was observed, again consistent with these models. The nucleation density measured by AFM increased from 2.7×107cm−2 at −75mVto2.5×109cm−2 at −300mV. Stress measurements with a combination of three additives [1×10−3moldm−3 Cl−, 8.82×10−5moldm−3 polyethylene glycol, and 1×10−5moldm−3 3-mercapto-1-propanesulfonic acid sodium salt (MPSA)] also showed that compressive stress generally developed initially and its magnitude was greater than in additive-free electrolyte. At less negative potentials, even though the rate of coalescence of nuclei was rapid, as observed by AFM, the stress continued to evolve in the compressive direction. At intermediate potentials (−90to−150mV), classical compressive-tensile-compressive (C-T-C) behavior was observed, while at more negative potentials the stress continued to evolve in the tensile direction. Similar results were obtained with a combination of two additives (1×10−3moldm−3 Cl− and 1×10−5moldm−3 MPSA), but in that case the compressive stress appeared to be greater, and consequently the T-C transition was observed even at −500mV. The results are consistent with enhancement of a compressive component of stress in the presence of additives.

A systematic investigation of the effect of annealing time and temperature on the incubation period for spontaneous morphology change ͑SMC͒ in electrodeposited copper metallization is reported. Based on an additivity principle derived by Mittemeijer for transformations with an Arrhenius-type dependence of kinetics on temperature, it is shown that the remaining incubation time for SMC at room temperature following annealing at a given temperature should be linearly related to the annealing time. By repeatedly scanning atomic-force microscopy images at room temperature, the time at which SMC occurred was determined for films annealed for various times at temperatures ranging from 32 to 60°C. At each temperature studied, the remaining incubation time at room temperature was found to decrease approximately linearly with increasing annealing time, thus experimentally verifying the behavior predicted by the additivity principle. An Arrhenius plot of the measured rates of decrease showed good linearity and yielded a value of 0.48 eV for the activation energy. This is consistent with a vacancy diffusion mechanism for the process occurring during the incubation period, and supports our proposed mechanism for SMC.Copper has been widely adopted as an interconnection material in ultralarge-scale integrated ͑ULSI͒ circuits. 1-3 Electrodeposition is the technology of choice for the fabrication of copper interconnects because of factors such as the ability to fill dual damascene architectures without voids, lower cost, and better electromigration performance. As a consequence of its rapid growth, the technology has in many respects outpaced fundamental scientific understanding, and many important questions need to be investigated if technological progress is to be maintained. Thus, the study of the deposition and properties of copper metallization films is of major importance to the semiconductor industry. In particular, understanding and controlling the microstructure of electrodeposited copper 4-22 metallization is increasingly important.It has been observed that the microstructure of both sputtered and electrodeposited copper can change considerably with time, even at room temperature. 4,23-33 The effects of additives to the plating bath on the electrodeposition and properties of copper metallization have been studied extensively. 15,18,19 Scanning probe microscopy, including atomic force microscopy ͑AFM͒ and scanning tunneling microscopy ͑STM͒, has proved to be a useful tool to study copper electrodeposition; [4][5][6][7][8][11][12][13][14][15][16][17][18][19][20][32][33][34][35][36][37][38][39][40][41][42][43] mechanistic studies, the effect of additives, and the influence of the substrate on microstructure have been reported. The influence of various additives on the roomtemperature recrystallization of electrodeposited copper has been investigated by Stafford et al. 22 using X-ray diffraction and resistance measurements. Lingk and Gross 27 demonstrated the influence of damascene topography and showed that initiation of r...

For Abstract see ChemInform Abstract in Full Text.