Dl. Vanderhart et al., Morphological characterization of blends of metal-sulfonated poly(styrene)and a methylated poly(amide) by solid state NMR, MACROMOLEC, 33(6), 2000, pp. 2206-2227
Various blends of atactic, low-MW (approximate to 4000), metal-sulfonated p
oly(styrene) (MSPS) and a higher-MW (approximate to 25 000) poly(amide) (PA
) were studied by solid-state C-13 and proton NMR techniques which include
multiple-pulse irradiation, cross-polarization, and magic angle spinning. T
his study is an investigation of the morphology of these MSPS(n)/PA blends
(n = 100 x mole fraction sulfonate = 2.3, 7.0, or 11.9) as functions of ble
nd composition and sulfonation level. Unsulfonated PS and PA are incompatib
le and phase separate. Decoration with sulfonate groups promotes mixing of
the blend components owing to strong, polar metal-sulfonate/amide interacti
ons, Metal ions used were divalent Zn (diamagnetic) and Cu (paramagnetic),
the latter ions having a significant influence on the protons. The PA, N,N'
-dimethylethylene sebacamide, was N-methylated to weaken interactions betwe
en PA chains, thereby promoting mixing. Pure PA is semicrystalline, and int
imate mixing prevents PA crystallization. C-13 CPMAS spectra were used to a
ssay PA crystallinity. The stability of the blend morphology in the presenc
e of water was also studied since water is expected to modify or compete wi
th the polar interactions of the blend. Many of the experiments performed r
elied, for their interpretation, on the phenomenon of proton spin diffusion
. For ZnSPS(11.9)/PA blends, mixing was quite intimate and PA crystallinity
was suppressed for PA mass fractions of 0.5 and lower. PA crystallinity fi
rst appeared with a PA mass fraction of 0.65; however, this crystallinity w
as not the result of large-scale phase separation of the PA from the MSPS.
Rather, PA crystallinity develops in the mixed MSPS/PA phase in such a way
that each PA crystallite is surrounded by a mixed MSPS/PA phase. Larger PA
mass fractions gave higher PA crystallinities. When PA crystallinity is pre
sent, there is an average periodicity of about 20-25 nm; moreover, the nonc
rystalline regions surrounding each crystallite have nonuniform composition
in the sense that there is a buffer zone adjacent to the PA crystallites w
hich is mainly PA in composition. A few blends involving ZnSPS(7.0) and ZnS
PS(2.3) were also studied. Only the 75/25 ZnSPS(7.0)/PA blend seemed well m
ixed and noncrystalline. Compositional heterogeneities on scales larger tha
n 20 nm were seen in the remaining blends. Certain CuSPS ionomers and blend
s, analogous to the Zn-containing materials, were studied in an attempt to
resolve some ambiguities present in the interpretation of the data taken fo
r the Zn-containing materials. Zn and Cu ions show similar affinities for t
he amide moieties; hence, the morphology for analogous Zn- and Cu-containin
g blends is expected to be similar. Mainly proton longitudinal relaxation w
as measured because it is sensitive to the presence of paramagnetic Cu. Two
matters were pursued for the Cu-containing materials: First, the uniformit
y of Cu distribution was probed in pure CuSPS(11.9). Our T-1(H) analysis ga
ve a variation by a factor of 1.3 in averaged Cu concentration, where the a
veraging was done over dimensions of 14 nm. Second. given that the 75/25 Cu
SPS(2.3)/PA blend exhibits large-scale phase separation, where one phase co
ntains nearly all of the PA plus a small fraction of the SPS, we addressed
the question whether the level of decorations for the SPS chains in this mi
xed phase was significantly above the 2.3% average. Our analysis did not su
pport such a claim within the assumption that the morphologies of the Zn- a
nd Cu-containing blends are the same.
Toward an evaluation of paramagnetic Cu ions as aids for the elucidation of
morphology in organic systems, several qualitative characteristics of the
proton and C-13 relaxation are noted for these Cu-containing materials. Als
o, an estimate of the electron relaxation time, T-1(e) in CuSPS(11.9) is gi
ven as is the fraction (0.95) of observable protons using multiple-pulse te
chniques. T-1(e) is shown to vary strongly with overall Cu concentration in
the pure ionomers and with the amount of water absorbed in the pure ionome
r. T-1(e) also changes when Cu is bound to the amide moiety of PA. One cann
ot simply assume that the average Cu concentration and (1/T-1(H)) are propo
rtional. A few annealing experiments also indicate that when the mole fract
ion of Cu is 7.0% or 2.3% in the ionomer, annealing seems to promote more c
lustering as though the cast films were not at their energy minima with res
pect to Cu-Cu interactions. Finally, in CuSPS(11.9), spin diffusion results
indicate that the position of the SAXS maximum near q = 2 nm(-1) is consis
tent with the separation between paramagnetic centers, which centers, in tu
rn, are estimated to consist of clusters of about 10 Cu sites. If cluster s
ize diminishes with average Cu concentration, then data indicate that T-1(e
) is a strong function of cluster size.