Dissolution behavior of poly vinyl alcohol in water and its effect on the physical morphologies of PLGA scaffolds

Article Information Received 15 January 2014 Received in revised form 19 Feb 2014 Accepted 20 Feb 2014 Abstract Presented are data from a study of the aqueous properties of Poly Vinyl Alcohol (PVA), a well studied emulsifying agent, used in the preparation of biodegradeable Poly (DL-Lactide-CoGlycolide) (PLGA) scaffolds/microparticles in water. How these properties affect the physical morphologies of PLGA scaffolds/microparticles produced from the various PVA solutions at different concentrations via the water emulsion synthetic method were also evaluated.


Introduction
Poly Vinyl Alcohol (PVA) shown in figure 1 below, is a white non toxic (included in the FDA inactive database) 1 , biodegradeable semi-crystalline polymer.It is produced via the hydrolysis of poly vinyl acetate under acidic/basic conditions 2 .It is also used as a stabilizing agent for emulsions (0.25-3.0%weight-per-volume), as a viscosity-enhancers particularly in ophthalmic products 1 .It is used in artificial tears and contact lens solutions for lubrication purposes 1 .

Made commercially available according to its degree of hydrolysis;
PVA can be classified into two main groups, the partially and full hydrolysed.The melting point varies depending on the degree of hydrolysis with the fully hydrolyzed grade having a melting point of ~228 °C, and the partially hydrolyzed grade ranging from 180-190 °C1 .
PVA has been reported to have good aqueous solubility due to its degree of polarity 20 but it also forms dilute polymer solutions that does not conform to the ideal-solution behavior [21][22][23] .Walker 24 explained that this could be due to a relational dependence between the Cohesive Energy Densities (CED) of both the solute and solvent, while small 25 summarized that the solubility of a polymer in UK Journal of Pharmaceutical and Biosciences Available at www.ukjpb.coma non-polymeric liquid depends solely on the heat of mixing.Tacx et al. 25 suggested that this must be due to ageing of the solutions leading to the formation of aggregates, a change they thought could be linked to the thermal history of PVA and the dissolution temperature.
Duda et al. [26][27][28][29] and Huggins 30 concluded that the molecular diffusion of a polymer in solution is a complex process, strongly dependent on temperature, concentration, polymer molecular weight as well as its morphology, they also reported the presence of an inter-molecular synergy between the polymer and the solvent molecules at elevated polymer concentrations which gives rise to a quicker dissolution of the polymer in solution.
These non-ideal behaviour of polymers in solution led Paul Flory 31 and Maurice Huggins 32 to develop a simple lattice model (Flory-Huggins model) that could be used to understand this unique characteristics of polymeric solutions.Based on a set of rules, the model in its simplest form 23  While the above generalisation applies only to cases involving low molecular weight solute, the entropy of mixing a higher molecular weight polymer is expected to be lower; this is due to a loss in conformational entropy brought about by the linkage of individual duplicative units along a polymer chain.Thus in expressing the ∆Sm for a higher molecular weight polymer in a solvent, the lattice is established by splitting the polymer chain into r number of segments, each size of a solvent molecule, where r is the ratio of polymer volume to solvent volume (in a lattice site).So N, the total number of lattice sites in this case (for n2 polymer molecules), is re-defined as N = n1 + rn2.

PVA solutions
Approximately weighed samples of PVA were added to 250ml of distilled water pre-heated to 40 °C, under continous stirring.The resulting solution is then stirred and heated (up to 65 °C) without any interruption until the PVA is completely dissolved.A total of ten samples of varying concentrations (w/v) were prepared; 0.1, 0.2......0.9,1.0 % (at higher concentrations PVA aggregation was observed in solution).

Absorbance measurements
This was taken using a Biocrim Libra S12 UV-spectrometer (Chemopharm).276 nm wavelength was used with distilled water as the reference sample.

Viscosity measurements
Solution viscosity was measured using the Brookfield viscometer (DV-I Prime); spindle 62 was deployed at a torque of 75%, with a 3 minutes acquisition time.

Scaffold preparation
The PLGA microparticles were prepared using the water emulsion synthetic method. 1 gm of PLGA was weigh and dissolved in 5ml of DCM, to this was added 250 µl of PBS solution.The resulting mixture was homogenized (Silverson L5M-A homogenizer; Fischer scientific, 25 Shah Alam, DE 40400, Malaysia) at 9000 rpm for 2 minutes.The new PLGA/DCM/PBS mixture was then added to 200ml of PVA solution which was then homogenized at 3000rpm for 2 minutes.The double emulsion was then stirred for 2 minutes at 300rpm and the microparticles formed was washed under continuous flow of water in a sieve (Fischerbrand test sieve number 230) and freeze dried.

Scanning Electron Microscopy
SEM images were obtained using the Phillip SL 30 (Koninklijke Philips Electronics N.V.) scanning electron microscope, at a voltage of 5KV.

Dissolution studies
As previously reported [21][22][23] , the dissolution of PVA in water over the entire concentration range used in the study was non ideal (in ideal situation, dissolution time is expected to increase as you have more solute in solution; as solubility decreases with increased concentration) as shown in figure 3 below; At low concentrations (0.1 -0.4), a linear increase in dissolution time with concentration was initially observed but after that (at 0.5 and 0.6), a very steep drop was seen, gradually increasing again at 0.7 and 0.8 before flattening out at 0.9 and 1.0 % w/v respectively.While we could not adduce this observation to any previous theorie(s), a more detail study into this abnormal behavior has currently been studied to see if such theories can be used to understand/explain it and would be published in a separate article.

Absorbance measurements
In order to determine if the amount of PVA weighed to make up the solutions used in the study correlated quantitatively with the actual amount of solute in solution, absorbance measurements was carried out on each of the solutions of PVA used.It is obvious from the figure 4 above that there is an increase in absorbance with the increase in concentration, suggesting that the amount of PVA in solution increases as the amount weighed to make up the solutions increases, we were unable to accurately estimate (within the limit of reasonable experimental error) the actual concentration in solution due to the non-complete linearity of the plot as the molar absorptivity of PVA could not be calculated accurately from the above calibration plot as defined by the Beer-Lambert's law 40 .

Viscosity measurements
The results obtained by measuring the liquid viscosities of the PVA solutions as a method of monitoring increase in concentration was calculated using the relationship in equation (c) 41 ; ƞsp = (ƞsolutionƞsolvent)/ ƞsolvent… (c) ƞsp = specific viscosity, ƞsolution = solution viscosity and ƞsolvent = solvent viscosity These were then plotted against concentration as shown in figure 5 below;

Fig 5: Plot of concentration against solution viscosity
Previous authors 42 concluded that the relationship between viscosity and concentration reaches a maximum first (in dilute solutions) and is then followed by a rapid decrease in viscosity due to the fact that the polymer (PVA) has reached its final stage of expansion, was not applicable therefore in this study.

Fig 1 :
Fig 1: Molecular structure of PVA First put into commercial use in Germany in the 1920s, PVA has wide applications in various industries: These include: textiles, paper, adhesives, cements, and films 3 .
considers the hypothetical mixing of a low molecular weight solvent 1 with a similar molecular weight solute 2 .Both solute and solvent molecules are assumed to have similar size, therefore a single lattice site can only be occupied by one solute or one solvent molecule at any given time.The increase in entropy (∆Sm) as a result of the mixing of both solvent and solute is then obtained using the Boltzmann relationship depicted in equation (a) below; ∆Sm = k ln Ω……………… (a) Given that k is Boltzmann's constant (1.38 x 10-23 J K-1), whereas Ω gives the total number of ways of arranging n1 identical solvent molecules as well as n2 identical solute molecules, where N = n1 + n2 is the total number of lattice sites while Ω the probability function is estimated as in equation (b) below; Ω = N!/(n1!n2!)....... (b)

Figure 2
Figure 2 below; a (lower molecular weight solute versus lower molecular weight solvent) and b (higher molecular weight polymer (single polymer chain) versus lower molecular weight solvent) respectively depicts the lattices in both cases.

Fig 2 :
Fig 2: Two dimension representation of the Flory-Huggins model ○ = solvent molecules, • = solute molecules Due to the limitations of the Flory-Huggings model, a new model was introduced in 1950, called the Flory-Krigbaum model[33][34][35][36][37][38] ; which presented new thermodynamic relationships of dilute solutions in which individual (heterogeneous) higher molecular weight polymer chains are isolated and at the same time encircled by regions of solvent molecules23 .Flory and Krigbaum considered the dilute solution as a distribution of clouds made up of polymer segments enclosed by regions of pure solvents as compared to the earlier model (Flory-Huggins); i.e. segmental density was no longer viewed as been uniform 23 .While previous authors have compared the dissolution properties of PVA and other polymers or different molecular weight PVA in various solvents, our study aims to look at the dissolution behavior of PVA in water only; by varying the concentrations of PVA (% w/v) added to a fixed volume of water to ascertain; (1) if the actual concentration of PVA in solution would increase quantitatively as the estimated concentration used, (2) how these changes in concentration would affect the physical characteristics of PLGA scaffolds produced from the differing PVA solutions.2 Materials and Methods PVA (fully hydrolyzed) was purchased from Sigma Aldrich, Co 3050 spruce street, St. Louis, MO 63103 USA.PLGA was provided by Evonik Degussa Corporation, Birmingham, AL 35211 USA.HPLC grade Dicloromethane (DCM) was obtained from Fischer scientific, bishop meadow road, Loughborough, UK, LE11 5RG.Distilled water was produced for this study using ELGA Purelab flex (Chemopharm, 47300 Petaling Jaya, Selangor, Malaysia), while Phosphate Buffered Saline (PBS) was supplied by Oxoid limted, Basingstoke, Hamshire, UK.

Figure 4 Fig 3 :Fig 4 :
Figure 4 below shows the plot of absorbance against concentration for different PVA concentrations;

Figure 6 4 Conclusions
Figure 6 below show some SEM images of the PLGA scaffolds prepared using the PVA solutions used in this study; The physical appearance/morphologies of the PLGA scaffolds were the same in all cases regardless of the concentration of PVA used in the studies; the scaffolds where spherical, porous with rough outer surfaces.4 Conclusions PVA exhibits a non-ideal solution behaviour in water, Nevertheless it was shown that irrespective of this, its concentration in solution increases as the amount measured and added to water increases (within the concentration range used in this study).The physical morphologies of PLGA scaffolds prepared from the PVA were not affected by the concentration of PVA in solution.