224
7 Composites
7.1.4 Cu(Ag) Alloys
Cu(Ag) composites gained much attention due to their prospective application in the
microelectronic industry as oxidation- and electromigration-resistant interconnect
material in which the favourably small resistivity of Cu can be retained. The target
range of the composition is < ~15 wt.% Ag. Cu–Ag deposits with a silver content
larger than 10% usually show segregation, which can be accompanied by the formation of a porous deposit [30]. Even if the deposit of high silver content is compact,
the X-ray diffractograms indicate [31] that the copper lattice cannot accommodate
the large excess of silver and the two metals are segregated.
Numerous bath types for obtaining Cu-rich Cu(Ag) alloys were proposed for
industrial application. Acidic baths were mostly formulated from the usual copper
sulphate bath containing sulphuric acid by adding AgNO 3 and used either without
[32] or with [33–35] some additives. Another acidic bath is based on methanesulphonic acid and methanesulphonate salts of both Ag and Cu [31]. One family of the
alkaline bath was based on cyanide compounds such as KCN, CuCN and K[Ag(CN) 2 ]
[36–38]. A cyanide-free alkaline bath with K 4 P 2 O 7 and KI has also been elaborated
[39, 40]. Regardless of the anion, the complexing agent and the additive used in
these baths, their common feature is that Ag
+ is discharged at more positive potential
than the Cu-containing species. The standard potential difference of the Ag
+ /Ag and
Cu
z+ /Cu (z = 1 or 2) systems is large enough so that even the complexation of the
cations does not modify the order of their reduction. Since the silver content of the
deposit is usually quite small, the Ag
+ reduction process takes place at the diffusionlimited rate. For the sake of easy regulation of the Cu:Ag ratio in the deposit, the
Ag
+ concentration in the bath is significantly smaller (0.1–50 mM) than that of the
couprous or coupric species, and the mole fraction of Ag in the deposit can be estimated from the ratio of the silver diffusion-limited current density to the total current
density.
The structural features of Cu(Ag) deposits can be easily elucidated by taking into
account the atomic volumes. As Ag atoms are larger than Cu atoms, the lattice is
expanded as the Ag is incorporated, and the lattice parameter measured increases
with the Ag content in the deposit [31]. This impact lasts to the Ag mole fraction at
which the segregation of Ag occurs, which manifests itself by the emergence of the
diffraction lines characteristic of pure Ag. Interestingly, as Ag forms a segregated
phase, the elimination of the nucleation barrier leads to a loss of the Ag content of
the metastable Cu matrix whose lattice parameter approaches again to that of pure
Cu [31, 36]. For the as-deposited samples, the Vegard law was found to be valid up
to an Ag concentration where a separate Ag phase is observed [4, 40].
Concerning the grain size of Cu(Ag) deposits, it has been shown [32] that the Ag
content itself has no major impact on the grain size of the deposit when the primary
variable is the Ag
+ concentration of the bath and the current density is constant.
However, the grain size decreases with increasing the current density [32, 33]. The
latter trend agrees well with those discussed in Chap. 2.11.
7 Composites
7.1.4 Cu(Ag) Alloys
Cu(Ag) composites gained much attention due to their prospective application in the
microelectronic industry as oxidation- and electromigration-resistant interconnect
material in which the favourably small resistivity of Cu can be retained. The target
range of the composition is < ~15 wt.% Ag. Cu–Ag deposits with a silver content
larger than 10% usually show segregation, which can be accompanied by the formation of a porous deposit [30]. Even if the deposit of high silver content is compact,
the X-ray diffractograms indicate [31] that the copper lattice cannot accommodate
the large excess of silver and the two metals are segregated.
Numerous bath types for obtaining Cu-rich Cu(Ag) alloys were proposed for
industrial application. Acidic baths were mostly formulated from the usual copper
sulphate bath containing sulphuric acid by adding AgNO 3 and used either without
[32] or with [33–35] some additives. Another acidic bath is based on methanesulphonic acid and methanesulphonate salts of both Ag and Cu [31]. One family of the
alkaline bath was based on cyanide compounds such as KCN, CuCN and K[Ag(CN) 2 ]
[36–38]. A cyanide-free alkaline bath with K 4 P 2 O 7 and KI has also been elaborated
[39, 40]. Regardless of the anion, the complexing agent and the additive used in
these baths, their common feature is that Ag
+ is discharged at more positive potential
than the Cu-containing species. The standard potential difference of the Ag
+ /Ag and
Cu
z+ /Cu (z = 1 or 2) systems is large enough so that even the complexation of the
cations does not modify the order of their reduction. Since the silver content of the
deposit is usually quite small, the Ag
+ reduction process takes place at the diffusionlimited rate. For the sake of easy regulation of the Cu:Ag ratio in the deposit, the
Ag
+ concentration in the bath is significantly smaller (0.1–50 mM) than that of the
couprous or coupric species, and the mole fraction of Ag in the deposit can be estimated from the ratio of the silver diffusion-limited current density to the total current
density.
The structural features of Cu(Ag) deposits can be easily elucidated by taking into
account the atomic volumes. As Ag atoms are larger than Cu atoms, the lattice is
expanded as the Ag is incorporated, and the lattice parameter measured increases
with the Ag content in the deposit [31]. This impact lasts to the Ag mole fraction at
which the segregation of Ag occurs, which manifests itself by the emergence of the
diffraction lines characteristic of pure Ag. Interestingly, as Ag forms a segregated
phase, the elimination of the nucleation barrier leads to a loss of the Ag content of
the metastable Cu matrix whose lattice parameter approaches again to that of pure
Cu [31, 36]. For the as-deposited samples, the Vegard law was found to be valid up
to an Ag concentration where a separate Ag phase is observed [4, 40].
Concerning the grain size of Cu(Ag) deposits, it has been shown [32] that the Ag
content itself has no major impact on the grain size of the deposit when the primary
variable is the Ag
+ concentration of the bath and the current density is constant.
However, the grain size decreases with increasing the current density [32, 33]. The
latter trend agrees well with those discussed in Chap. 2.11.