THESIS
2013
xxii, 164 pages : illustrations ; 30 cm
Abstract
The heart of three-dimensional (3D) Si integration is the copper filled Through Silicon
Via (TSV), which allows the shortest chip-to-chip interconnections. The copper filling of
the TSV is usually achieved with an electro-plating method. Even though copper
electro-plating for interconnects is a well-known technology due to the wide application of
the copper damascene process, it proves to be quite different from the copper filling of the
TSV, where the via diameter changes from nanometers to tens of microns and via depth
changes from sub microns to hundreds of microns. What we have learned from the copper
damascene process and what works there could not be applied directly to TSV copper filling.
Although some void-free TSVs have been achieved, there are still many technical issu...[
Read more ]
The heart of three-dimensional (3D) Si integration is the copper filled Through Silicon
Via (TSV), which allows the shortest chip-to-chip interconnections. The copper filling of
the TSV is usually achieved with an electro-plating method. Even though copper
electro-plating for interconnects is a well-known technology due to the wide application of
the copper damascene process, it proves to be quite different from the copper filling of the
TSV, where the via diameter changes from nanometers to tens of microns and via depth
changes from sub microns to hundreds of microns. What we have learned from the copper
damascene process and what works there could not be applied directly to TSV copper filling.
Although some void-free TSVs have been achieved, there are still many technical issues to
be solved for the copper filling process, including low throughput, high cost ( accounts for
nearly 40% of the overall TSV cost), heavy overburden and inconsistent void-free filling.
Copper electro-plating is traditionally conducted utilizing copper sulfate based electrolyte
containing three different functional additives. Recently a new acidic copper methane
sulfonate based electrolyte has been promoted as one candidate copper plating bath.
However, literature survey shows that less research work has been performed on the copper
electrodeposition in the TSV filling process. It is important and necessary to understand the electrochemical performance of the new sulfonate based electrolyte, TSV filling capability
and compatibility with the existing additives.
This thesis is devoted to implementing a comprehensive study of the performance
characteristics of the new copper methane sulfonate based electrolyte and its compatibility
with the existing additives. The ultimate goal is to screen and develop an additive system
that is compatible with sulfonate based electrolytes and which can successfully fill most of
the vias without any voids. A number of experimental methods such as wetting-ability tests,
Hull-cell studies, current efficiency analysis, polarization behavior studies, electrochemical
evaluation and TSV copper filling were carried out to investigate the performance
characteristics of the new electrolyte.
Performances of three industrial additives were firstly investigated based on the
sulfonate electrolyte. The sulfonate bath is compatible with the industrial additives and
void-free copper filling is achieved for 55/220 μm (diameter/depth) vias, utilizing the
sulfonate bath containing industrial additives within 6 hours. However, void-free copper
filling cannot be realized for vias of 25 μm in diameter and 200 μm in depth due to poor
wetting-ability. The polyethylene glycol (PEG)/ Bis-(3-Sulfopropyl)-disulfide sodium salt
(SPS) system was then screened and involved in the sulfonate electrolyte. The
wetting-ability of the sulfonate bath may be improved by the addition of a SPS accelerator,
while the incorporation of PEG suppressor has a minor influence on the wetting
performance. In addition, void-free copper filling is accomplished for the 55/220 μm vias
using the electrolyte containing PEG/SPS within 4 hours. However, the system cannot
successfully fill the vias of 25 μm in diameter and 200 μm in depth.
Finally, a single additive system was developed and successfully applied in the copper
filling of three different vias with the diameters of 55, 25 and 10μm. In addition, this single
additive copper filling process may simplify the additive replenishment and monitoring
procedure during the TSV filling process. The correlation of the copper filling of TSV and
cyclic voltammetry striping (CVS) monitoring of the plating bath was built up to predict the
formation of voids inside the vias. The proposed void-free copper filling criteria may be
further applied in other plating electrolytes. Finally, wafer-level TSV copper filling was
successfully achieved utilizing the single additive system.
Post a Comment