When polymer solution is subjected
to ionizing radiation, reactive intermediates are formed.
This can result from direct action of radiation on the polymer
chains and from indirect effect, i.e. reaction of the intermediates
generated in water with polymer molecules (for a general description
of radiation-induced processes in aqueous solution, see e.g.
[43]). Since the fraction of energy absorbed by each component
of the polymer-water system is proportional to its electron
fraction, which can be well approximated by the weight fraction,
in dilute and moderately concentrated polymer solutions the
indirect effect dominates. The input of the two effects to
the yield of polymer radicals is usually even more shifted
to the indirect effect than it results from the weight fraction,
since the yield of radicals in water is, in general, higher
than in pure polymer itself. Therefore the description below
will refer to the indirect mechanism only.
Fig. 5. Pulse radiolysis of vinylpyrrolidone
solution, monomer concentration -0.94 M. Oscilloscope traces
recorded during light scattering detection at two different
doses, pulse length - 1 ms,
solution saturated with N2O.
Out of the three main reactive species formed in water upon
irradiation - hydrated electrons, hydroxyl radicals and hydrogen
atoms - electrons exhibit low reactivity towards simple, hydrophilic
hydrogel-forming polymers. This is an expected behavior (found
also for low-molecular-weight models of these polymers), since
they usually do not contain functional groups being efficient
scavengers of hydrated electrons. Rate constants of these
reactions can be estimated by pulse radiolysis technique,
by following the changes in the lifetime of hydrated electron
with increasing polymer concentration. The values of the rate
constant for these reactions are usually lower than 1*107
dm3 mol-1s-1
[12].
Hydroxyl radicals have been shown to be the main species responsible
for reactivity transfer from water to the polymer chains.
They abstract hydrogen atoms from macromolecules, thus polymer
radicals are formed. Pulse radiolysis allows to determine
the rate constants of these reactions, either by following
the increase in the absorbance of macroradicals being formed,
or, what is usually more convenient, by the competition method.
Detailed studies have shown that for polymers the reaction
with OH cannot be described with a single value of rate constant,
since such a rate depends on polymer concentration and molecular
weight [64-66]. In general, the rate constants, calculated
based on the molar concentration of the monomer units (and
not the chains), decrease with the increase of the chain length
and they are close to the diffusion-controlled limit (for
discussion see e.g. [12,66]).
From the data available for small organic molecules similar
to our polymers and also from the results obtained e.g. for
PAA at low pH (where e-aq
are converted to H) one can conclude that hydrogen atoms react
with these molecules in a similar way as OH radicals, i.e.
by abstracting hydrogen atoms, however the corresponding rate
constants are expected to be somehow lower.
Only in rare cases, like poly(ethylene oxide), all the hydrogen
atoms in the polymer molecule are equivalent, so that only
one type of macroradicals can be formed upon hydrogen abstraction
by OH radicals. Otherwise hydroxyl radicals abstract hydrogen
atoms from various non-equivalent positions, so that two or
more kinds of radicals of different structure are formed (Fig.
6, cf. [45,57,58,67-69]). For the polymers listed above, these
macroradicals are localized on carbon atoms. Structure of
these carbon-centered polymer radicals may be of some importance
for the process of crosslinking and hydrogel formation. Some
macroradicals may be more prone to undergo side reactions,
and even if the same type of reaction is considered - for
example chain scission, it may either lead to the breakage
of the main polymer chain, when the radical was localized
there, or to some minor chemical changes in the side groups
only, when the radical was localized on a side chain [70].
Since the structure of crosslinks and the energy of interchain
bonds depends on the structure of the recombining radicals,
one could anticipate that the localization of radicals before
the crosslinks are formed may influence the structure and
properties of the final product - hydrogel.
The fractions of radicals formed at various positions can
be estimated in a number of ways - analysis of absorption
spectra of transient products after an electron pulse, selective
oxidation of the radicals and EPR. Straightforward EPR measurements
of the radical spectra are rarely made, because of the rapid
decay of these transients. Only in exceptional cases, like
ionized PAA, radical lifetimes are long enough to enable recording
of the EPR spectra [58,59]. An alternative method is the spin-trapping,
allowing to transfer the initial macroradicals into the radicals
of high stability (cf. [58]). One can also irradiate a low-molecular-weight
model, identify and quantify the products, and, by analyzing
the possible reaction ways, estimate the initial fractions
of radicals at various positions.
On the basis of data available for PVAL, PVP and PAA one can
conclude that the selectivity of OH attack on these macromolecules
is not very high, i.e. there are no positions that would be
the sole target of an OH attack (cf. Fig.2 in [11]). In the
case of PAA the fractions of radicals in the a-
and b-positions to the carboxy
group (ca. 30% and 70 %, respectively) are similar to the
fractions of H-atoms available for abstraction from these
positions [58]. Some selectivity was observed for PVAL, where,
as for simple alcohols [71], attack at the a-carbons
to the hydroxy groups is preferred [45]. Also in PVP two of
the possible five positions are the main attack sites [57].
Fig. 6. Structures of polymer radicals
formed upon irradiation of simple water-soluble polymers in
aqueous solution [45,57].
From the practical point of view - formation of hydrogels
- the most important reaction of macroradicals is intermolecular
crosslinking. i.e. recombination of radicals localized on
two different macromolecules (Fig. 7a). Although other various
reactions in discussed systems are known since long [72],
their importance and role in the competition with intermolecular
crosslinking is not always fully recognized. If only the latter
reaction would occur in irradiated aqueous polymer solution,
one could expect the yield of intermolecular crosslinks, Gx,
defined as number of crosslinks formed in the system upon
absorption of 1 Joule of ionizing radiation energy, to be
equal to the half of the initial OH and H yield. This value
would then be Gx = 1.6*10-7
mol/J for deoxygenated (i.e. saturated with Ar or N2)
solutions and ca. 3.0*10-7 mol/J
for solutions saturated with nitrous oxide (where the OH yield
is doubled, see e.g. [43]). In fact the values of Gx
are often much lower (for examples see Table 1 in [11]), what
indicates that many of the initially formed macroradicals
undergo other reactions. These other reactions include reactions
between two radicals, as intramolecular crosslinking as well
as inter- and intramolecular disproportionation, and also
processes involving one radical, as hydrogen transfer or chain
scission. These processes do not result in joining the polymer
chains, thus they do not lead to the formation of macroscopic
gels.
The proportion between recombination and disproportionation
reactions is set by the radical structure and the possibility
to control this parameter is usually very limited. What is
important, model studies for PVAL and PAA show that in these
systems disproportionation is the main reaction involving
two radicals, while the fraction of recombination is ca. 20
- 35 % for PAA model and only ca. 10 % for the PVAL model
[45,58]. Thus the overall yields of crosslinking (both inter-
and intramolecular) in the corresponding polymers are limited
to these values.
We do have, however, an influence on the competition between
inter- and intramolecular crosslinking. At high polymer concentration
(above the critical hydrodynamic concentration, which depends
on the molecular weight), when polymer chains interpenetrate,
the probability that two recombining radicals are localized
on different chains is relatively high. What's more, in such
systems some physical entanglements may become "fixed"
when the entangled chains become joined to the network in
at least two points encompassing the entanglement site. If
we lower the polymer concentration to a range where the macromolecules
(usually having a conformation of a coil) are separate, then
the probability of intermolecular recombination decreases.
If the irradiation conditions allow for the simultaneous existence
of more than one radical on a single chain, then, provided
the chain is flexible, the encounters of two radicals on the
same chain may become faster than intermolecular recombination
requiring two big chemical entities as polymer coils to diffuse
to each other (Fig. 7b). Besides concentration, another parameter
of equal importance for this competition is the way of irradiation,
or, more precisely, the dose rate. High dose rates, as achieved
by pulse-irradiation of the system with electron beam, when
combined with low polymer concentration, may lead to the situation
when several tens of even more than a hundred radicals are
generated simultaneously on each chain. In these conditions
the probability and yield of intermolecular recombination
is greatly reduced. These effects have been studied in many
polymeric systems (cf. in aqueous solution of star-shaped
poly(ethylene oxide, [73]).
The influence of polymer concentration on the competition
between inter- and intramolecular crosslinking emerges from
the plots of the gelation dose vs. concentration. A typical
curve illustrating such relationship (Fig. 8, first examples
were shown by Alexander and Charlesby [72]) consists of two
parts. On the high-concentration side the gelation dose is
almost proportional to the polymer concentration. This is
in line with the general theory of crosslinking that assumes
that a gel is formed when, on average, one crosslink is formed
for one macromolecule present in the system. However, when
the concentration is lowered beyond some limiting value, instead
of a linear decrease there is a pronounced increase in the
gelation dose. This is equivalent to a strong decrease in
the yield of intermolecular crosslinking. This is the concentration
range where intramolecular recombination prevails. The dose
rate effect has been demonstrated e.g. by Ulanski et al. [45]
by pulse radiolysis with low-angle light scattering detection.
When a sample of N2O-saturated PVAL
solution has been subjected to a single electron pulse of
150 Gy, almost no increase in the scattered light was detected.
However, when the same dose was administered by a series of
low-dose pulses, a significant increase could be observed,
indicating the increase in molecular weight as a result of
intermolecular crosslinking.
The competition of two recombination modes can be followed
by kinetic studies as well. While intermolecular reactions
follow the classical second-order kinetics, intramolecular
recombination usually shows significant deviations from this
simple kinetic pattern [45,58,75,76]. It can be shown that
in the latter case the reaction rate depends on the average
number of radicals per chain, rather than on the overall radical
concentration. For quantitative description of the reaction
rates, a model of non-homogeneous kinetics has been successfully
applied [77,78]. For further details as well as an attempt
to explain the reasons for the non-classical kinetic behaviour
- see [66]. To summarize, high polymer concentration and low
dose rate promotes gel formation, while for dilute solutions
irradiated with high dose rates one can expect the dominance
of intramolecular recombination.
Fig. 7. Schematic representation
of (a) intermolecular crosslinking due to the fixation of
entanglements in concentrated polymer solution and (b) intramolecular
crosslinking in diluted solution of isolated chains. Dots
on macromolecule before reaction denote radical sites.
Hydrogen transfer reactions change the location of radical
sites, but do not change the overall number of radicals on
polymer chains, so that in general their occurrence does not
decrease the number of radicals available for crosslinking.
Nevertheless, some influence on the crosslinking yield and
network structure can be expected, by changing the initial
proportions between various radical structures that may be
more or less prone to recombination and, as already mentioned
above, may lead to various gel microstructures.
Chain scission, being in a sense a reverse process to intermolecular
crosslinking, is an important reaction for our discussion.
In cases when it proceeds with high yield, exceeding Gx, no
gel formation occurs in the system. In deoxygenated solutions,
when the chain break precursors are carbon-centered radicals
localized at the main chain (or its immediate vicinity), chain
scission reactions are, fortunately, very slow. Thus, in most
cases of the polymer considered here (polyelectrolytes being
an exception), radicals recombine before chain scission proceeds
to a measurable extent. Certainly, an increase in the molecular
weight of the polymer, or even gel formation do not prove
the absence of degradation. However, sol-gel studies as well
as analysis of the irradiation products of the model compounds
allow to estimate the scission yield. For most non-ionic polymers,
like polyacrylamide, polyvinylpyrrolidone and poly(vinyl alcohol),
these yields, under standard irradiation conditions, were
found to be close to zero. However, this is no longer true
for ionic polymers and in the case of the presence of oxygen
in the system.
Figure 8. Exemplary dependence of the
gelation dose on the polymer concentration. Data for polyvinylpyrrolidone
K-30, g-irradiated in an Ar-saturated aqueous solution. Dose
rate 3.0 kGy/h [79].
Poly(acrylic acid) requires a separate paragraph, as an exemplary
polyelectrolyte. Its radiation-induced transformations in
solution has been recently studied in some detail [58,59,74,75,80].
PAA can be effectively crosslinked by irradiation with no
need for any additives [81], provided the irradiation is performed
in acidic solution (e.g. pH 2) so that most of the carboxylate
groups are protonated. Under these conditions PAA resembles
a non-ionic polymer and its behavior under irradiation can
be described (although still not fully) by the general rules
applicable for uncharged macromolecules. Increasing pH towards
neutral and alkaline reveals the specific properties of PAA
as a polyelectrolyte. High density of negative charge on the
chain, only in part screened by the condensed counterion atmosphere,
induces coulombic repulsive forces between the chain segments
and the macromolecule assumes a rod-like, relatively stiff
conformation. When radicals are generated on the macromolecules,
these forces prevent the radical-bearing chain segments and
also the neighboring chains from an approach into the reaction
distance, thus slowing down recombination and disproportionation
by seven orders of magnitude in comparison with uncharged
chains. Since the radical-terminating reactions are that slow
(in this system macroradicals can live for hours at room temperature),
competing reactions as hydrogen shift and, what’s more important
here, chain scission proceed with high yield and wins the
competition against crosslinking. As a result of that, the
molecular weight of the polymer decreases (the net yield of
chain breaks exceeds 5*10-7 mol/
J for 1*10-2 mol/dm3 PAA at pH >
9) and no gel is formed.
The chain scission processes are even more important for poly(methacrylic
acid) (PMAA) [70,82,83]. The macroradicals of PMAA, which
are very long-lived especially at high pH, undergo b-scission
reaction (1), giving end-chain radical and unsaturated terminal
structural element.
The rate constant of this reaction depends significantly on
the acidity of the medium and it is highest in the range of
pH 7-9. Additional process taking place in the irradiated
PMAA solution is very effective chain unzipping (2/3) i. e.
depolymerization.
As a result of this process, methacrylic acid is formed with
high yield (at pH 9 a dose rate of 0.09 Gy/s and a dose of
20 Gy G(methacrylic acid)=500*10-7
mol/J was measured) [82,83]. The rate of unzipping reaction
depends strongly on the pH, and it is only very prominent
at high pH. It is worth noting that the rate of scission for
the dissociated PMAA radicals is ca. 70 times faster than
that of PAA radicals under similar conditions [58]. This much
slower rate of fragmentation is paralleled by the fact that
an unzipping process (reaction (2/3)) was not observed with
the PAA-derived radicals. The differences in the chemical
behavior between these two polyelectrolytes can be caused
by various radical stabilization and chemical structure. In
PMAA the most stable radicals in the a-position
to the carboxylic group cannot be formed due to the polymer
chemical structure. Formed b-radicals
are less stable and easily undergo the transformation via
scission processes. Moreover, possible radical recombination
for PMAA is hindered from sterical reasons. Additionally,
some influence of the hydrophobic effects on the chain conformation,
operative for poly(methacrylic acid), cannot be ruled out.
PAA hydrogels formed by irradiation of its acidic solutions
or by radiation copolymerization with crosslinking agents
belong to the class of stimuli-sensitive materials and respond
by swelling or shrinking to the changes in pH, ionic strength
and electric field [15,84].
Since PMAA undergoes chain scission with a considerable rate
constant and even depolymerize rapidly at high pH, it is impossible
to crosslink poly(methacrylic acid) under these conditions
by ionizing radiation. In principle, this could be done at
low pH and at high dose rate delivered by electron beam, when
the radical lifetime is shorter and the rate of scission slower.
However, at such very high dose rates intramolecular crosslinking
are promoted rather than macroscopic gel formation (see above).
In order to favor the intermolecular process high concentration
of PMAA is required. However, at a 10 % solution, some small
particles are formed, but a wall-to-wall gel was not yet observed
[83].
|
