branch cells usually manifest covalent connectivity relative to some molecular
reference marker (I) or core. As such, these branch cell arrays may be very
non-ideal and polydispersed (e.g. M w /M n ffi 2–10), as observed for random
hyperbranched polymers (IVa), or very ideally organized into highly controlled
core–shell type structures, as noted for dendrons/dendrimers (IVc) (M w /M n ffi
1.01–1.0001 and less). Dendrigraft (arborescent) polymers reside between these two
extremes of structure control, frequently manifesting rather narrow polydispersities of
M w /M n ffi 1.1–1.5, depending on their mode of preparation.
3.3 Dendritic Polymer Subclasses
3.3.1 Random Hyperbranched Polymers
Flory first hypothesized dendritic polymer concepts [32, 30], which are now
recognized to apply to statistical or random hyperbranched polymers. However,
the first experimental confirmation of dendritic topologies did not produce random
hyperbranched polymers but rather the more precise, structure-controlled,
dendrimer architecture [43, 44, 46, 55]. This work was initiated nearly a decade
before the first examples of random hyperbranched polymers were confirmed
independently by Gunatillake, Odian et al. [57], as well as by and by Kim and
Webster [58, 59] in 1988. At that time, Kim and Webster coined the popular term
“hyperbranched polymers” that has been widely used to describe this subclass of
dendritic macromolecules. Hyperbranched polymers are typically prepared by
polymerization of AB x monomers. When x is 2 or more, polymerization gives
highly branched random polymers, as long as A reacts only with B from another
molecule. Reactions between A and B from the same molecule result in termination
of polymerization by cyclization. This approach produces hyperbranched polymers
with a degree of polymerization n, possessing one unreacted A functional group and
[(x – 1) n + 1] unreacted B terminal groups. In a similar fashion, copolymerization
of A 2 and B 3 or other such polyvalent monomers can give hyperbranched polymers
[60, 61] if the polymerization is maintained below the gel point by manipulating
monomer stoichiometry or limiting polymer conversion. Random hyperbranched
polymers are generally produced by the one-pot polymerization of AB x -type monomers or macromonomers involving polycondensation, ring opening, or
polyaddition reactions. Hence, the products usually have broad, statistical molecular weight distributions, much as observed for traditional polymers. Over the past
decade, literally dozens of new AB 2 -type monomers have been reported, leading to
an enormously diverse array of hyperbranched structures. Some general types
include poly(phenylenes) obtained by the Suzuki coupling [58, 59]; poly
(phenylacetylenes) prepared by the Heck reaction [62]; polycarbosilanes, polycarbosiloxanes [63], and poly(siloxysilanes) by hydrosilylation [64]; poly(ether ketones)
by nucleophilic aromatic substitution [65]; and polyesters [66] or polyethers [67] by
polycondensations or by ring-opening polymerization [68].
336
D.A. Tomalia
reference marker (I) or core. As such, these branch cell arrays may be very
non-ideal and polydispersed (e.g. M w /M n ffi 2–10), as observed for random
hyperbranched polymers (IVa), or very ideally organized into highly controlled
core–shell type structures, as noted for dendrons/dendrimers (IVc) (M w /M n ffi
1.01–1.0001 and less). Dendrigraft (arborescent) polymers reside between these two
extremes of structure control, frequently manifesting rather narrow polydispersities of
M w /M n ffi 1.1–1.5, depending on their mode of preparation.
3.3 Dendritic Polymer Subclasses
3.3.1 Random Hyperbranched Polymers
Flory first hypothesized dendritic polymer concepts [32, 30], which are now
recognized to apply to statistical or random hyperbranched polymers. However,
the first experimental confirmation of dendritic topologies did not produce random
hyperbranched polymers but rather the more precise, structure-controlled,
dendrimer architecture [43, 44, 46, 55]. This work was initiated nearly a decade
before the first examples of random hyperbranched polymers were confirmed
independently by Gunatillake, Odian et al. [57], as well as by and by Kim and
Webster [58, 59] in 1988. At that time, Kim and Webster coined the popular term
“hyperbranched polymers” that has been widely used to describe this subclass of
dendritic macromolecules. Hyperbranched polymers are typically prepared by
polymerization of AB x monomers. When x is 2 or more, polymerization gives
highly branched random polymers, as long as A reacts only with B from another
molecule. Reactions between A and B from the same molecule result in termination
of polymerization by cyclization. This approach produces hyperbranched polymers
with a degree of polymerization n, possessing one unreacted A functional group and
[(x – 1) n + 1] unreacted B terminal groups. In a similar fashion, copolymerization
of A 2 and B 3 or other such polyvalent monomers can give hyperbranched polymers
[60, 61] if the polymerization is maintained below the gel point by manipulating
monomer stoichiometry or limiting polymer conversion. Random hyperbranched
polymers are generally produced by the one-pot polymerization of AB x -type monomers or macromonomers involving polycondensation, ring opening, or
polyaddition reactions. Hence, the products usually have broad, statistical molecular weight distributions, much as observed for traditional polymers. Over the past
decade, literally dozens of new AB 2 -type monomers have been reported, leading to
an enormously diverse array of hyperbranched structures. Some general types
include poly(phenylenes) obtained by the Suzuki coupling [58, 59]; poly
(phenylacetylenes) prepared by the Heck reaction [62]; polycarbosilanes, polycarbosiloxanes [63], and poly(siloxysilanes) by hydrosilylation [64]; poly(ether ketones)
by nucleophilic aromatic substitution [65]; and polyesters [66] or polyethers [67] by
polycondensations or by ring-opening polymerization [68].
336
D.A. Tomalia