Chemistry Reference
In-Depth Information
such as polyimide (PI), poly(etherimide) (PEI) and poly(tetrafluoroethylene)
(PTFE) and have shown that the increase of the Pdac concentration leads to an in-
crease of the Pd nuclei density, and finally to an increase of the practical adhesion
of the Cu deposit (peel test).
3.3. Adhesion characteristics of electroless films
The last part of the present work aimed at investigating the adhesion characteris-
tics of Ni thin films electrolessly deposited on Kapton
®
500 HN substrates. In all
the cases which were previously qualified as providing bright and adherent Ni de-
posits, the use of the cross-cut peel test showed that no metal part was removed
by the Scotch
®
tape, and therefore that a good practical adhesion had been ob-
tained between the metal and the polymer. This also means that the effects of the
different efficient treatments used to prepare the PI surface before metallization
cannot be distinguished through this test.
For a similar purpose, a fragmentation test has also been employed. It should
be pointed out that such a test has already been used for studying both fracture
and delamination of metallic or ceramic thin, brittle films adhering to ductile sub-
strates [60-65]. Usually, the coated specimens are strained under uniaxial tension
and the development of parallel cracks which are transverse to the straining direc-
tion is observed (by electron or optical microscopy) as the applied strain pro-
gresses. According to the theoretical approach by Wheeler and Osaki [60], the
maximum interfacial shear stress as it is inferred from experimental data (crack
density at saturation) can be taken as a parameter related to the interfacial adhe-
sion. In the present experiments, an original approach has been employed [33, 40,
41]. It consists in subjecting the coated specimens (in dog-bone shape) to the uni-
axial tension and measuring the resulting changes in the electrical resistance of
the metal films between two metallic contacts symmetrically located at the ends
of the narrow part of the investigated specimens. As the main objective of these
experiments is to show the possible effects of surface treatments on the practical
adhesion of the metal films to their substrate, it is necessary, for a comparison
purpose, to deposit Ni films of same composition, structure and thickness. Conse-
quently, such films must be deposited from the same plating bath (here the com-
mercial Europlate Ni 520 solution) and their thickness determined. This means
that calibration curves must be established. As an example, Figure 9 illustrates the
dependence of X-ray fluorescence data (Ni K
α
intensity measurements) on the ini-
tial conductance (1/R
o
) of the Ni films. In the thickness range investigated, the Ni
K
α
intensity and film initial conductance vary linearly with the film thickness. As
a result, a mere measurement of the initial electrical resistance R
o
of the Ni films
provide access to their thickness.
Coming back to the fragmentation test itself, Figure 10 is plotted for 200 nm
thick Ni films deposited on PI substrates which were previously subjected to dif-
ferent surface treatments. The (a) and (b) curves show how the relative electrical
resistance
δ
R/R varies as a function of the applied strain, i.e. in fact against the
Search WWH ::
Custom Search