Grain Growth, Stress, and Impurities in Electroplated Copper
- PDF / 448,469 Bytes
- 8 Pages / 612 x 792 pts (letter) Page_size
- 48 Downloads / 237 Views
E. Kerr IMEC, Kapeldreef 75, B-3001 Leuven, Belgium, and Science of Materials, Trinity College, Dublin, Ireland
I. Vervoort IMEC, Kapeldreef 75, B-3001 Leuven, Belgium
A. Saerens Department of Metallurgy and Materials Engineering, K.U.-Leuven, B-3001 Leuven, Belgium
K. Maex IMEC, Kapeldreef 75, B-3001 Leuven, Belgium, and Electrical Engineering Department, K.U.-Leuven, B-3001 Leuven, Belgium (Received 5 February 2000; accepted 21 December 2001)
The widely observed secondary grain growth in electroplated Copper layers is shown to be incomplete after the sheet resistance and stress of the layer appear to have stabilized. Instead the layer is in an intermediate state with a grain size distribution that depends on the plating conditions. Further extensive annealing at high temperatures results in an additional considerable enlargement of the grain structure, accompanied by an additional decrease of the sheet resistance and desorption of impurities that were incorporated during plating.
I. INTRODUCTION
The introduction of copper as the new interconnect material of choice has necessitated a firm understanding of the physical properties of this material in both thin films and narrow structures (trenches). It quickly became clear that electroplating provided the most advantageous means of deposition in terms of cost and, equally important, filling of trenches without voids.1 The latter was achieved through the use of several additives in the electroplating bath that result in an accelerated growth from the bottom of the trench upward.2 Although the use of these additives facilitates an easy and effective way to fill high aspect ratio structures, it also induces two less desirable phenomena. Firstly, the accelerated ‘bottom-up’ growth tends to overshoot, producing large hillocks over patterned areas.3 Secondly, one of the additives (known as brightener) also causes a reduction in as-deposited grain size to a subcritical value. This induces a phenomenon known as self-annealing, in which secondary grain growth occurs at room temperature over a period of days or weeks, accompanied by stress changes and a 20% decrease in sheet resistance RS. In our early work we showed that sheet resistance and stress are in fact due to two different phenomena.4 Sheet resistance changes in copper thin films have been shown to relate to changes in grain structure, while the asdeposited stress varies with impurity concentration. Focused ion beam microscopy (FIB) shows a close 582
J. Mater. Res., Vol. 17, No. 3, Mar 2002
correlation between secondary grain growth and the decrease of RS. Some discrepancies between the two have been reported,5 but we believe that these are due to the fact that RS is sensitive to both normal and secondary growth, while FIB reflects the latter only. Through the brightener concentration [B], we could vary the as-deposited stress 0 over a range of tensile values (80 > 0 > 30 MPa).6 Because all of these stress levels reduce to near-zero values during selfannealing, grain boundary volume elimination could be rul
Data Loading...
