Biocompatible and Biodegradable Ultrafine Nanoparticles of Poly ( Methyl Methacrylate-co-Methacrylic Acid ) Prepared via Semicontinuous Heterophase Polymerization : Kinetics and Product Characterization

Ultrafine nanoparticles, less than 10 nm in mean diameter, of the FDA approved copolymer methyl methacrylate(MMA-) comethacrylic acid (MAA), 2/1 (mol/mol), were prepared. The method used for the preparation of these particles stabilized in a latex containing around 11% solids includes the dosing of the monomers mixture on a micellar solution preserving monomer starved conditions. It is thought that the operation at these conditions combined with the hydrophilicity of MMA and MAA units favors the formation of ultrafine particles; the propagation reaction carried out within so small compartments renders copolymer chains rich in syndiotactic units very likely as consequence of the restricted movements of the end propagation of the chains. Because of their biocompatibility and biodegradability as well as their extremely small size these nanoparticles could be used as vehicles for improved drug delivery in the treatment of chronic-degenerative diseases.


Introduction
The copolymer methyl methacrylate-(MMA-) co-methacrylic acid (MAA), 2/1 (mol/mol), is a biodegradable material long time approved by the FDA, which is used for the elaboration of drug-loaded pills because of its water solubility at pH > 7 [1].This copolymer is marketed under the commercial name of Eudragit S-100 (ES-100) [1].Many literature studies have documented the preparation of drug-loaded micro-and nanoparticles making use of ES-100.The processes involved usually include the dissolution of a drug-copolymer mixture in a suitable solvent which is subsequently evaporated to finally obtain the drug-copolymer particles.These particles usually show mean diameters in the order of hundred nanometers [2][3][4][5][6][7][8][9][10][11]; the exception is the work of Dai et al. [3] who obtained drug-loaded nanoparticles with 37 nm in mean diameter, as measured by quasi-elastic light scattering (QLS).It is noticeable that none of these processes include the polymerization step.In fact, reports on preparation of poly(MMA-co-MAA) 2/1 mol/mol nanoparticles via polymerization were not found in our literature search.This is at odds with the increased potential of smaller particles as drug carriers, which could be produced by an appropriate polymerization method.In accordance with literature, drugloaded polymeric nanoparticles with diameters of less than 50 nm and, more specifically, ranging 10-30 nm are very attractive.They would show the ability to cross through the intestine wall and enter the blood stream [12]; because of their smallness, their ability to circumvent the immunological system would be increased allowing its circulation in the 2 International Journal of Polymer Science blood stream for longer periods of time [13].Moreover, nanoparticles smaller than 50 nm in diameter are more easily internalized by cells [13].
Recently we have reported the obtaining of a latex containing poly(MMA-co-MAA), 2/1 mol/mol ultrafine nanoparticles loaded with Ibuprofen showing 9.2 nm in mean diameter, determined by scanning-transmission electron microscopy (STEM) [14].The solid content in the latex was 12.8%, which is relatively high for a dispersion containing particles so small.In that document we described the method named as semicontinuous heterophase polymerization at monomer starved conditions for latex preparation along an explanation for obtaining such small particles.This method was first reported some years ago [15,16] and basically consists in dosing the monomer at a low enough flow on a micellar solution containing the appropriated amount of surfactant to stabilize the small particles generated.As an additional interesting feature, micelles are absent at the end of the polymerization.
In this report we document a research on the polymerization to prepare poly(MMA-co-MAA), 2/1 mol/mol nanoparticles; the polymerizations were conducted following the recipe and under the conditions used in our previous work [14].Herein is described the kinetic behavior as well as the evolution of particle size, copolymer composition, and elution times of the copolymer in gel permeation chromatography with polymerization time.

Polymerization.
It was conducted in a 150 mL jacketed glass reactor equipped with a reflux condenser and mechanical agitation using the following procedure.93 g water, 0.1 g APS, and a 1.0 g mixture of surfactants SDS/AOT, 3/1 (wt/wt), were charged into the reactor, after which the mixture was subjected to 650 rpm agitation and the temperature stabilized at 70 ∘ C. The reaction was initiated when the addition of the monomers mixture (12.6 g of MMA/MAA, 2/1 (mol/mol)) was started; the addition was carried out at constant flow for one hour using a dosing pump (Kd Scientific-100).At the end of the dosing period the polymerization was allowed to continue for 30 min.Due to the relatively large sample amounts required for the characterizations, a series of runs was carried out for the polymerization.For this, the addition period was divided into 5 intervals (12,24,36,48, and 60 min); then a run for each interval was carried out, stopping the polymerization at the end of each of them and collecting all the latex to carry out characterizations.A run including the 30 min postaddition period was also performed.With this protocol, the amount of latex obtained was enough to make all the required characterizations.

Monomer Conversion.
The conversion values of monomers to polymer along the polymerization were determined by gravimetry; for this purpose, the latex sample was dried to constant weight in an oven at 60 ∘ C. Then the weights of the surfactant and initiator contained in the sample were subtracted to that of the dried sample in order to know the amount of polymeric material produced at a given time.

Particle Size Determinations.
Measurements by QLS were conducted in a Malvern Zetasizer Nano-ZS90 apparatus at 25 ∘ C. To eliminate multiple scattering and particle interactions, the latexes samples were diluted 50 times with water.The size distribution and average sizes of the particles in the latexes were determined from measurements in a JEOL JSM-7401F field emission scanning electron microscope (FE-SEM) operated at scanning-transmission mode (STEM) using a transmission electron detector SM74230RTD.The samples were prepared by mixing one latex drop with 10 g of water.
Then, one drop of this dispersion was deposited on a copper grid and allowed to dry.Once dried, the sample was stained using osmium tetroxide vapors.The diameters of more than 100 particles were measured from the micrographs by using the image analysis program ImageJ1.37c.From these data   ,   , and PDI (  /  ),   and   being the weightand number-average diameters, respectively, and PDI the polydispersity index, were calculated using the following equations [17]: where   is the number of particles of size   and  is the total number of measured particles.

Surface Tension.
This property was measured at 25 ∘ C in a SensaDyne PC500-L tensiometer.

Nuclear Magnetic Resonance (NMR) Spectroscopy.
The copolymer samples were analyzed by 13 C (12,000 scans) NMR in a Bruker-400 MHz spectrometer.A mixture of deuterated chloroform and deuterated trifluoroacetic acid (50 : 50 volume ratio) was used as solvent and the analyses were performed at room temperature.

Thermal Analysis.
The determination of the thermal behavior of the polymer was carried out in a modulated differential scanning calorimeter (DSC) TA Instruments Discovery Series DSC.Measurements were performed at a heating rate of 10 ∘ C/min in the range 0 to 275 ∘ C, under a nitrogen flow of 50 mL/min.After the first heating scan the samples were quenched and a second scan was carried out.The glass transition temperatures (  ) were evaluated from the second scan.

Gel Permeation Chromatography (GPC) Analysis.
For these determinations the surfactant was previously eliminated from the latex by dialysis in water using porous membranes (Sigma-Aldrich) with exclusion size corresponding to molecular weights larger than 12,500 g/mol.The dialyzed dispersion was then subjected to three freezingand-thaw cycles for polymer coagulation.The copolymer free of surfactant was recovered by filtration, washed with deionized water, and dried, before dissolution in HPLC-grade tetrahydrofuran (THF) (Merck).An Agilent PL-GPC50 gel permeation chromatograph equipped with three columns (Agilent PLgel) and a refractive index detector was employed, using THF as the mobile phase.The equipment was calibrated with polystyrene (PS) standards (Polyscience) covering the molar mass range of 6.0 × 10 2 to 3.0 × 10 6 g/mol.

Results and Discussion
3.1.Kinetics.Polymerization evolution with time is shown in Figure 1, where two kinds of conversions can be distinguished: instantaneous (  ) and global () conversion.The former is defined as the fraction of added monomers up to time  that has converted into polymer, while the latter represents the fraction of total monomer in the formulation converted into polymer at time .As can be seen in this figure,   attains values close to 90% no later than 24 min since the monomers addition was started.Taking into account that the particles are formed of a polymer-monomers mixture and that a small fraction of the monomers would be swelling the micelles, this means that the remaining monomers in the system represent around 0.1 weight fraction at best, which is a value much less than the monomers saturation concentration.This assumption arises from the fact that MMA shows a saturation concentration of 0.71 volume fraction [18] and that the corresponding value for MAA, a monomer much more soluble in water than MMA, should be even higher because in accordance with literature the saturation concentration values increase as the monomer water solubility increases [18].The consequences of polymerizing at conditions in which the monomer concentration in the particles is below its saturation value are very interesting.This particular operating mode is known as monomer starved conditions, whose foundations were first reported by Krackeler and Naidus [19], which in turn is based on the emulsion theory by Smith-Ewart [20].This theory includes a correlation stating that the particle number in latex ( p ) is inversely proportional to the rate of volume increase () of the particles during nucleation period.One way for minimizing  is getting very low monomer concentration within the particles, that is to say, operating at monomer starved conditions; this operating mode can be achieved by dosing the monomers at a very low flow on the initial micellar solution [15,16].As described above, this would maximize  p and, as a consequence, it could allow very small polymer particles to be obtained.

Particle Size.
values determined by QLS for nanoparticles at different conversions are shown in Table 1, which also includes   and PDI values from STEM measurements for a low and final conversion. Figure 2 shows the particle size distributions from QLS measurements at different polymerization stages.On the other hand, micrographs from latex samples at low and final conversions along their corresponding histograms of particle size are shown in Figure 3. Data in Table 1 indicate that average particle size grows at the initial stage of polymerization attaining values of 15.5 nm as determined by QLS; then it decreases and finally stabilizes around 10 nm as the polymerization evolves.This behavior suggests that a competing generation of smaller particles than those already produced commences before reaching the polymerization half point, which leads to a reduction in the average particle size.In accordance with the theoretical basis described above on polymerization at monomer starved conditions, this effect could be due to a decrease in the polymerization rate within the particles, which in turn would cause a decrease in .To probe this hypothesis a calculation of global polymerization rate ( p ) was performed using the  versus  data in Figure 1 along the polymerization recipe; the obtained  p values are also included in Table 1.Additionally,  p (1/L of water) at different conversions was calculated using the following equation: where  p is the polymer concentration in g/L of water, calculated by using the corresponding ,  p , the copolymer density, taken as 1.2 g/mL [21], and   , the number average diameter determined by QLS, in cm. p values shown in Table 1 indicate that particle nucleation exists practically through all the reaction, which is a well-known fact in semicontinuous heterophase polymerization at monomer starved conditions with high enough surfactant concentrations [16,22,23].In our case this condition was favored not only by the relatively high surfactant concentration but also by the contribution as stabilizers of the MMA and mainly of MAA units, because of their hydrophilic character [24].However, the decrease in  p at the postaddition period suggests a certain degree of particle coalescence.Once the values of  p and  p during the polymerization are known, the polymerization rate per particle,  pp , defined as  p / p was calculated for different  values, showing the corresponding results in Table 1.Undoubtedly,  pp decreases as polymerization evolves; this decrease is more marked at the first stage of the reaction, precisely when the average particle size diminution is observed.This  pp behavior supports the hypothesis stated above which explains the behavior of average particle size with the polymerization evolution.
Another interesting point, in this case related to the possible use of this type of particles as drug carriers, should be mentioned.It refers to the absence of free surfactant forming micelles once the polymerization is completed.This is deduced from the surface tension () behavior of the latex during the polymerization, shown in Figure 4, where it can be seen that  remains less than 38-39 mN/m at  < 90%, that is to say, during the monomers dosing period.At the end of this period,  starts to grow attaining a final value of 43.2 mN/m, suggesting that all the surfactant in the formulation has been used in the particle stabilization [25].To confirm this, an estimation of the particle surface covered by all the surfactant contained in the formulation was carried out, using the following calculation sequence.From  p in Table 1 the total particle number in the final latex was calculated; with this value and the surface area of a particle the total particle surface was known ( tp ).Then, the total particle area able to be covered by the surfactant molecules in the formulation ( ts ) was calculated.For the latter calculation the surfactant content in the formulation and the values of its weighted molecular weight (316.2 g/mol) were used.
The surfactant content in the formulation was used for the latter calculation, as well as the values of its weighted molecular weight (316.2 g/mol) and the particle area that a surfactant molecule covers in a saturated monolayer (0.55 nm 2 ) [26].Thus, the results obtained were 5.9 × 10 21 and 1.1 × 10 21 nm 2 for  tp and  ts , respectively, which led to a  ts / tp ratio of 0.19.This means that all surfactant molecules in the formulation are only about one-fifth of those required to form a saturated monolayer on the particles, which suggests that there is no residual surfactant to form micelles.Nevertheless, despite this partial coverage, the latex has remained stable after more than five months, showing a   value close to 10 nm.It is believed that the polar groups pending in the copolymer chains, that is to say, carboxylic acid and ester groups, contribute to the stabilization of the particles by acting as cosurfactants.
As it has been shown, this method allows obtaining latexes composed of polymeric particles with mean diameters close to 10 nm containing around 11% solids.

Copolymer Composition.
According with the literature on the copolymerization of MMA (monomer 1) and MAA (monomer 2), the values of the reactivity ratio of MAA ( 2 ) are in the range from 0.2 to 0.4, while those corresponding to the reactivity ratio of MMA ( 1 ) are approximately the double of these amounts [27,28].In a batch copolymerization of these monomers, at the molar ratio used in this work, their reactivity ratios would lead to a mixture of moderate alternating copolymer [29] with homopolymer of MMA.However, in this study, the 13 C-NMR results for samples of polymeric material collected at different global conversions allow obtaining MMA/MAA molar ratios of 2.2 (Figure 5(a)) and 2.1 (Figure 5(b)) at 8.0 and 89.7%, respectively.These values indicate that a copolymer with MMA/MAA molar ratio similar to that in the feed was obtained.These MMA/MAA molar ratios were calculated using the signals of the carbonyl groups, which are located between 178 and 182 ppm for the MMA and between 182 and 184 ppm for the MAA.Moreover, DSC results for polymeric material at the same conversions support the conclusions from 13 C-NMR analysis.This can be seen in Figure 6, where only one   appears in the 25 to 250 ∘ C interval, with values ranging from 149.8 to 156.4.These values are somewhat different to that reported, 130 ± 5 ∘ C, as the   value for the copolymer MMA-co-MAA, 2/1, mol/mol [30].It should be mentioned that the copolymerizations carried out under semicontinuous operation at monomer starved conditions in an aqueous medium stabilized by surfactants allow the synthesis of homogeneous copolymers having a composition similar to that of the dosing mixture of monomers [31,32].This stems from the fact that the incorporation to the copolymer chain of one or the other monomer is mainly dictated by their relative availability within the particles rather than by their reactivity ratios.In a given moment, the consumption of the more reactive monomer causes its depletion within the particle allowing an increase of the chain incorporation ratio of the less reactive one; the situation prevails until the relative concentration of the former attains values that again enable its preferential incorporation.This interchange leads to the generation of an alternating copolymer during the monomers mixture dosing period, avoiding the formation of MMA homopolymer; once this period has finished, the copolymerization would proceed as in a batch fashion; however, the amount of polymeric material formed in the postaddition period is very low, so the effect on its final composition is negligible.
Since the MMA/MAA molar ratio in the copolymer is close to 2/1, the question arising is why the   values obtained along the polymerization are up to 25 ∘ C higher than that reported in the literature.A possible answer is the reported fact that syndiotacticity is favored when MMA polymerization is carried out in microemulsion [33][34][35] [36].In these quoted works PMMA with syndiotactic contents of more than 60% were obtained [33], which resulted in   values around 130 ∘ C, which is higher than the 115 ∘ C value reported for commercial PMMA [37].This is wellmatched with the   range between 130 and 141 ∘ C expected for pure syndiotactic PMMA [38].That behavior is explained by Jiang et al. as a result of the polymeric chain growth in the restricted space provided by the nanoparticles [34].Under this condition the end propagation of the chains with high molecular weight was also restricted during the polymerization and as the energy of the racemic dyads was lower than that of the meso dyads, it could result in a configuration rich in syndiotactic content [34].Figure 5(c) shows a part of the 13 C-NMR spectrum of the polymeric material collected at 89.7% global conversion, in which the signals between 179 and 181 ppm correspond to those of C=O in a predominantly syndiotactic PMMA [36,39].Despite the fact that the polymeric material prepared in this work is not PMMA but a copolymer MMA-co-MAA, its high content of MMA units, or more specifically predominantly syndiotactic MMA units, leads us to conclude that the copolymer obtained is one with an important content of syndiotactic units, which in turn is the cause for the higher   values obtained.
3.4.GPC Analysis.Figure 7 shows the elution time for copolymer samples obtained at  of 28.7 and 79.3%.The shifting toward the lower values of the elution time distribution curve at the higher conversion suggests that the polymer chains at the final stage of the polymerization are longer than those at the initial one.Despite the fact that the molecular weights of the samples were not determined because of the unavailability of the Mark-Houwink constants for the copolymer, it is obvious that, in a similar way, the molecular weights at the end of the polymerization would be larger than those at the initial part.To explain this behavior a discussion on the termination mode of the growing chains within the particles should be included.Chain growing termination occurs by chain transfer or by bimolecular reactions [39]; on one hand, chain transfer can take place to an agent such as the initiator or the surfactant in the medium; moreover it can occur to the monomers or the polymer.On the other hand, bimolecular termination would occur between the growing chain within the particle and an oligomeric radical entering into it.Chain transfer reactions could be discarded because of the following arguments.First, to the best of our knowledge, chain transfer International Journal of Polymer Science 7 to APS initiator as well as to SDS and AOT surfactants has not been reported in the polymerization of MMA or MAA in aqueous media.This does not imply the impossibility of their occurrence, but the probability should be very low.Second, chain transfer to monomer would render a molecular weight invariant during the polymerization, since polymerization degree does not depend on monomer concentration [40]; this possibility was discarded in the light of the behavior of elution time distribution in Figure 7. Third, chain transfer to polymer would lead to a significant increase of molecular weight, or a much marked diminution in the elution time with conversion, resulting from the formation of branched chains [40], which was not observed by us.Furthermore, to the best of our knowledge, this latter chain termination mode in polymerizations of MMA or MAA carried out in aqueous media stabilized with surfactants has not been reported.Thus, the remaining option of those mentioned above, bimolecular reaction, would probably be the termination mode of chains within the particles in our study.The difference in the elution times observed in Figure 7 appears to support this statement.The shorter elution times, indicative of larger molecular weights, at the end of the polymerization imply an extension of the growing time of the copolymer chain, which would be a consequence of a slowdown in the bimolecular termination rate between the chain and the entering radical into the particles.This condition would have its origin in a decrease in the radical entering rate due to a diminution in the ratio of radicals in the aqueous medium to particles, since   greatly increases as polymerization evolves (see Table 1).

Conclusions
The copolymerization of MMA/MAA, 2/1 (mol/mol) was carried out at semicontinuous operation and monomer starved conditions in an aqueous medium stabilized by surfactants.Under these conditions a latex with 11% solids content was obtained, containing particles close to 10 nm mean diameter, formed of a copolymer whose composition equals the feed monomer ratio used.It is believed that the operation at monomer starved conditions combined with the hydrophilicity of MMA and MAA units favors the formation of ultrafine particles.In turn the propagation reaction within so small compartments gives copolymer chains rich in syndiotactic units.To the best of our knowledge, this is the first report on the preparation of a latex with such small particles of the copolymer in question.Given that this copolymer is approved by FDA our method provides an option for preparing drugloaded nanoparticles for the improved treatment of chronicdegenerative diseases, taking into account the additional advantages offered by so small particles.

Figure 5 :Figure 6 :
Figure 5: 13C-NMR spectra of samples from the copolymerization of MMA/MAA (2/1, mol/mol) taken at global conversions of (a) 8.0 and (b) 89.7%.The segment of the spectrum of the sample at 89.7% in global conversion at which the syndiotacticity is evidenced is also included (c).
a determined by QLS; b determined by STEM.