The importance of the film structure during self-powered ibuprofen salicylate drug release from polypyrrole electrodeposited on AZ31 Mg

Therapeutic drugs uploaded into conjugated conductive polymer matrices deposited on active magnesium alloys serve as controlled-dose, self-powered drug-delivery systems. Preferentially, drugs are added into polymer films in the largest amount possible, mostly to prevent long-term treatments. However, added drugs can interact with the polymer matrix affecting either the structure or the final mechanical properties of the polymer film. In this work, polypyrrole films (PPy) electrodeposited on an AZ31 Mg alloy in ibuprofen and salicylate-containing solutions are investigated in terms of their uploading capacity, surface morphology and mechanical properties. The techniques used to investigate the uploaded PPy films include cyclic voltammetry (CV), scanning electron microscopy (SEM), EDS, and depth-sensing indentation (DSI). A maximum ibuprofen concentration of 440 ± 40 μg cm−2 was obtained in PPy films in the presence of sodium salicylate. The release fraction of ibuprofen as a function of time is fitted to Avrami’s equation. The hardness and reduced modulus decreased by 54 and 40 %, respectively, when the PPy films are prepared in the presence of sodium ibuprofen compared with those prepared in sodium salicylate only, indicating a more plastic film with ibuprofen.


Introduction
Polymer films electrodeposited on active metal alloys and uploaded with various ionic therapeutic drugs are the subject of current investigations [1][2][3]. Some of them are encouraged by the development of biodegradable, biodisposable self-powered drug-delivery systems [1].
The operation mechanism of self-powered systems relies on a galvanic coupling between conductive polymers films deposited on the surface of an active alloy. Therapeutic drugs are uploaded as dopants into the film, while the polymer is growing by attracting counterions to balance the charge [4][5][6]. Once a self-powered system is implanted in a corrosive environment, as the one encountered inside the human body, or on the surface of the skin, it undergoes corrosion. The corroding active alloy pumps electrons to reduce the polymer, contracts and releases the therapeutic drug. Accordingly, a sufficient corrosion resistance alloy act as a biodegradable implant and in combination with a polymer film serve as vehicle to deliver drugs such as risperidone, dexamethasone, pain relievers or antibiotics [1,4].
Typical alloys for self-powered drug release include, Mg (97% Mg, 2% Al, 1% Zn) [1], Mg (99.9% purity) [2], sputtered Mg thin films [5] and AZ91D and AZ31 Mg alloys [7,8]. These alloys are attractive as they match the mechanical properties of the human bones. For instance, Mg alloys and the human bone have an approximate density of 1.74-2.00 g cm -3 and 1.8-2.1 g cm -3 respectively, and an elastic modulus of 41-45 GPa and 130-180 GPa, respectively. Moreover the compressive yield strength of Mg is 65-100 GPa while the human bones is 3-20 GPa [9]. One concern however, is the corrosion of biodegradable Mg alloys that can release undesirable metals such as Al, so biocompatible alloys having a low Al content, such as AZ31 Mg alloy, are desirable for biomedical applications [10,11].
Polypyrrole is one of the most common, biocompatible and mechanically robust conductive polymers (CPs) which has been used for corrosion protection of metals [12,13,7,14] and local drug delivery [4]. The size, steric effects, functional groups, charge and chemical structure of the uploaded drug may affect the polymer nucleation and growth processes and its mechanical properties [13]. The depth-sensing indentation (DSI) technique can be used to determine the mean hardness (H) and reduced modulus (Er) of PPy films electrodeposited under different circumstances [15][16][17].
During the electrodeposition of the conductive polymer film on active magnesium alloys, two competitive processes occur; the nucleation and growth processes of the polymer and the dissolution of the metal. Dopant ions such as salicylate and oxalate form a passive layer on the Mg surface and allow the polymerisation [18,19]. However, when drugs are present in solution, they might be absorbed on the surface of the active magnesium alloys affecting the CPs electrodeposition [20]. It has been reported in the literature that ibuprofen affects the polymerisation and structure of poly(3,4-ethylenedioxythiophene) (PEDOT) and poly (3,4-ethylenedioxypyrrole) PEDOP matrices [6,21]. Thus, the amount of drug incorporated into CPs needs to be optimized [22]. Several alternatives have been suggested to increase the amount of the drug on the polymer such as nanostructure CPs [22] or addition of cationic surfactants [23] or heparin that increases the cation exchange capacity [24].
The aim of this paper is to investigate the effects of sodium ibuprofen in the morphological structure, mass fraction release rate and mechanical properties of electrodeposited PPy films on AZ31 Mg alloys in 0.5 mol dm -3 sodium salicylate solution at different IBU concentrations, 0.5-50 x10 -6 mol dm -3 . The novelty of this paper is to highlight the effects of ibuprofen on the quality and properties of PPy films deposited on AZ31-Mg alloy.

Experimental
Pyrrole monomer (Sigma Aldrich) was double-distilled and stored in a dark flask at low temperature, under a nitrogen atmosphere. Sodium ibuprofen, (Fluka), sodium salicylate (Aldrich) and acetone (Fisher Scientific) were analytical grade and used as received. All The concentrations of IBU and SA were determined spectrophotometrically at wavelengths of 222 and 295 nm, respectively, using calibration curves. Since SA shows absorption at 222 nm wavelength, it was necessary to subtract its absorbance at this wavelength in order to determine the IBU concentration. The reported data correspond to an average of at least three experiments. The PPy films were characterised by with a Nicolet 360 FTIR spectrophotometer and the spectra compared with spectra reported elsewhere in the literature [7,8].

Electrochemical characterisation
Open-circuit potential (OCP) measurements of the AZ31 Mg alloy in SA-Py, and SA-Py to that in SA solutions (curve a), but for concentrations above 10 × 10 -3 mol dm -3 IBU, the OCP becomes much more noble reaching -1.26 V vs. SCE after 600 s as shown in Fig. 1f) for 50 × 10 -3 mol dm -3 IBU. This occurs because passive hydroxide/oxide and salicylate magnesium chelate films, which can further diminish the magnesium dissolution, contribute to the equilibrium at the alloy/solution interface [18,19]. No oscillations, indicating breakdown and reconstruction of the passive layer, were seen during the OCP measurements.
The increase of the OCP with increasing concentration suggests that the IBU contributes to the passivation of the magnesium surface. Before growing a stable compact PPy film on the surface of an AZ31 Mg alloy, it is critical to leave the sample at the OCP for at least 600 s in the SA solution to passivate the surface, otherwise an incomplete PPy film will be formed.
Cyclic voltammograms (CVs) of an AZ31 Mg alloy in three different electrolytes are shown in Fig. 2. In an SA solution, curve (a), a large oxidation peak at ca. -0.35 V vs. SCE due to the oxidation of a non-passivated Mg alloy, which is far away from the OCP (-1.62 V vs. SCE) can be seen. Mg ions react with salicylate in solution forming a low solubility chelate complex that precipitates on the metal surface [18]. At +0.8 V vs. SCE an extended second anodic process is due to the oxidation of salicylate ions in solution [8]. This process is irreversible because there is no reduction peak. The analysis of the oxidation peak at -0.35 V vs. SCE at different potential sweep rates shows that the peak does not follow a linear relationship with the scan rate nor with the square root of the scan rate in the range of 10-70 mV s -1 . This suggest that the oxidation is neither a diffusion-controlled process nor an adsorption, instead, it confirms that the formation of the passivating chelate layer on the alloy limits diffusion of magnesium ions and forms a stable magnesium chelate complex. The peak shifts towards less negative potentials as the scan rate increases, which is consistent with an irreversible process [25].
The CVs responses for solutions containing SA-Py (curve b) and SA-Py-0.5IBU (curve c) in SCE and lower current density compared to the SA solution (curve a) due to formation of the salicylate complex. In these solutions, nucleation and growth of the PPy film at a potential at ca. +0.8 V vs. SCE takes place [8,25]. The nucleation and growth processes are evidenced by the crossing potentials at +0.8 V and +0.73 V vs. SCE for SA-Py and SA-Py-0.5IBU solutions respectively, during the reverse scan towards negative values [8]. At more positive values than the PPy nucleation and crossing potentials, there is a rapid increase in the current due to the polymerisation of Py and the dissolution of the underlying AZ31 Mg alloy. The two solutions seem to follow a similar trend, but a higher current density was observed for SA-Py and a more blunt reversal current wave for SA-Py-0.5IBU, indicating that less PPy film was formed on the electrode in the presence of IBU.

Influence of concentrations IBU
Figs. 3a and 3b show a series of CVs acquired in solutions with higher IBU concentrations, i.e., 10 and 50 × 10 -3 mol dm -3 , respectively. The first cycles for both concentrations do not present the characteristic peak due to the formation of the Mg salicylate film and the current density, due to the polymerisation of PPy is very low specially during the first cycles, at more than +0.8 V vs. SCE, as seen in Fig. 2. This suggests that IBU not only affects the formation of the passive salicylate adsorbed layer, but also the nucleation and growth of the PPy film.
As the number of cycles increases the solution with 10  10 -3 mol dm -3 IBU shows a slight increase in current density due to the polymerisation of Py at > +0.8 V vs. SCE. However, the PPy film does not form a uniform coating on the AZ31 Mg alloy surface. Instead, a series of randomly distributed black PPy spots with poor adherence and several cracks were formed, which readily detached from the surface revealing underlying corrosion products. Figure 3b shows that an increase in the IBU concentration up to 50 × 10 -3 mol dm -3 the formation of the PPy film is completely inhibited. Similar results have been reported in the literature on the effect of IBU on the electropolymerisation of PEDOP-IBU films, where the polymerization is suppressed at high IBU concentrations and films with lower charge capacity are formed [21]. This effect has been explained in terms of a transition between the ibuprofen acting as a dopant and becoming a non-dopant inclusion [21]. In this case, the salicylate layer was absent and the current density was substantially reduced without any potential crossings that denote the PPy nucleation and growth. It can be speculated that IBU is adsorbed in the Mg alloy limiting the access of Py to the electrode surface. In addition, the incorporation of IBU into the polymer structure, either as a dopant or inclusion, also limits the growth processes of the polymer likely due to steric effects. IBU is a large molecule in comparison with other counter ions, such as Clor ClO4used to balance the charge during the polymerisation and changes nucleation and growth processes [26,27,21]. cm 2 area of the AZ31 Mg alloy was 3.0 × 10 -3 g. The highest load is 414 × 10 -6 g/cm 2 of IBU for sample 4 which correspond to a ratio of 138 × 10 -3 g g -1 (IBU/PPy). This value is within the range of concentration of dopants in PPy films which varies from a few × 10 -9 g to several × 10 -3 g, either for nanostructured or chemically assisted drug uploading in PPy films [1,23].
On the other hand, IBU released after 5 min elapsed-time from a 1 cm 2 drug-loaded matrix from PEDOP-IBU system is between 21.8 ±1.3 to 25±1.3 x10 -6 g cm -2, more than one order of magnitude smaller than in this work, for PPy-IBU-SS films [21]. These large differences are likely due to a more substantial inclusion of IBU in the film besides the charge storage capacity and expected differences between the release efficiency of both films. Other examples of the amount of drug than can be uploaded on PPy films include neurotrophin-3 with a value of 25 × 10 -9 g cm -2 on gold coated Mylar at a PPy film of 3.6-26 × 10 -6 m thickness [22]. Another one is methotrexate on PPy films with a loading capacity of 24.5 × 10 -3 g g -1 [23] and chlorpromazine with 191±7 × 10 -3 g on a 0.126 cm 2 Pt disc coated with 8±2 × 10 -6 m thick PPy film [24].
IBU was released from 20±5 x10 -6 m PPy film deposited on one side of 1 cm 2 AZ31 Mg alloy immersed in a 0.9% NaCl solution. The other side of the alloy was bared and exposed to the NaCl solution. Fig. 6 shows a release profile curve of mass fraction of drug released as a function of time. At short periods of time (<50 minutes), the rate of release is high and tends to level off. This result is consistent with previous reports in the literature denoting a rapid release at the initial stages for dexamethasone in 10% wt. NaCl solutions and other systems [1,4,28]. Self-powered systems rely on the dissolution of the substrate metal to reduce the polymer film and release the incorporated drug. Any through porosity or cracks in the PPy film can also provide a pathway for chloride ions to diffuse through to the substrate metal and promote crevice and pitting corrosion in the PPy/AZ31 Mg alloy interface [28,29]. This increases the corroding area above the fixed area at the back of the sample. Moreover, it also accounts for a large scattering in the data for the release fraction at larger periods of time, where the error bars enlarge substantially likely due to the unpredictable corrosion rate (see Fig. 6 for t >250 min). This is consistent with previous results reported in the literature which shows large scattering at longer times, for self-powered systems [1].
Avrami's equation (1) was used as a model to represent the fraction release profile X for samples prepared using a concentration of 0.5 × 10 -3 mol dm -3 IBU. This equation was initially proposed to analyse and predict crystallization processes, but it can also fit well to diffusion controlled and potential-assisted drug release data, particularly for PPy in salicylate media and different types of drugs [30,31,24]. Both Avrami's parameter, n, and the release constant, k, are obtained from the slope and intercept of the Avrami's plot of ln[-ln(1-X)] vs.
According to the literature, for diffuse controlled drug release systems, the Avrami's parameter, n, varies from 0.54 to 1 (first order), depending on the contribution of the diffusion [30,31,24]. In the case of self-powered release, the constant, k, is equal to 0.125, which is within the reported values for potential-assisted release processes of ionic drugs, but the Avrami parameter, n, kinetic order parameter, is equal to 0.11, much smaller than values reported in the literature [31,24]. Besides these differences, Avrami's equation describes the trend of the experimental data obtained for the self-powered IBU release from PPy films.
The contribution of potential difference and diffusion in self-powered release systems can be expected due to potential variations across the film [4].

Depth-sensing indentation (DSI)
The mechanical properties of the PPy films are dependent on the chemical structure, molecular size and dopant ion concentration [17], [15]. To investigate the effect of ibuprofen in the PPy film, a series of 20 indentations along randomly selected sampling lines were performed on IBU-free and IBU loaded PPy films. Fig. 7 shows load-displacement curves obtained from PPy films uploaded with different amounts of ibuprofen. For each sample the 3 × 10 -3 N load resulted in a different depth of penetration, with the maximum depth of the indents in the 5 × 10 -3 mol dm -3 doped sample being less than 1µm, the indentation interaction volume being confined to the PPy film and not influenced by the substrate. Table 1 summarizes the average hardness (H), reduced modulus (Er) and the ratio between H and Er of the films investigated, as well as the concentration of IBU in solution and contained in the PPy film. Comparison of the mechanical properties of ibuprofen-free PPy films with those containing large amounts of ibuprofen (414 × 10 -6 g), showed a decrease in the H and Er ratio of 54% and 40%, respectively. The (H/Er) increases 66% denoting a softer more elastic material. Theoretically, the addition of ibuprofen into the PPy film increases the free space in the polymer matrix and modifies the structure of the PPy film [15,17] sodium ibuprofen is a large molecule with a single charge, when added into the PPy matrix acts as an additional ion dopant along with the sodium salicylate ions to balance the polymer charge.
The size of the ibuprofen molecule however results in an increase in the gap between the polymer chains, reducing the strength of the attractive forces between them making polymer chain motion and deformation easier than IBU-free polymer films [15,17]. As previously shown vide supra, in the absence of ibuprofen, PPy films in SA media developed a uniform, compact, adherent, globular-like structure (Fig. 4a)-b)). In contrast, the addition of ibuprofen into the PPy film produces a less compact, less adherent, cylindrical, needle-like structure in Fig. 4c)-d). There are few reports in the literature concerning the role that dopant ions play in PPy in terms of their effects on molecular conformations, charge location and structural defects in the polymer chain [32,33]. Usually, ion dopants induce changes in the polymer conformation that result in either a better alignment or increased misalignment of the polymer chains and contribute to additional packaging or unpacking of the polymer matrix [32]. The mechanical properties of the PPy films can be also modified by the electrosynthesis method used to prepare the film or by additional heat treatments [34]. This provides the opportunity to prepare films which are optimized in terms of their elasticity for specific purposes, such as a more corrosion protective films or more reliable drug delivery systems.

Conclusions
The effects of sodium ibuprofen on the quality of the electrodeposited PPy films were characterised. A high loading IBU capacity in the PPy films is desirable for long-term therapeutical treatments. The amount of IBU incorporated into the PPy film in this work is higher than that reported in the literature for other drugs or IBU loaded in PEDOP and PEDOT matrices. However, additions of IBU to the solution affect the electrodeposition processes of the film. A transition from a globular to cylindrical morphology in a PPy film has been observed for IBU concentrations in solution >1  10 -3 mol dm -3 . Larger IBU concentrations of 50  10 -3 mol dm -3 completely inhibit formation of the PPy film. Avrami's equation provides an appropriate fit to the experimental IBU fraction released from a PPy film in 0.9 wt. % NaCl solutions. IBU also produces substantial changes in the mechanical properties of the PPy film compared with IBU-free PPy films. The results indicate up to 54% reduction in H, 40% reduction in Er, and an increase of 66% of H/Er. An increment of the polymer free volume is the major cause for diminishing the mechanical properties of the PPy film. The outcomes from this investigation require the modelling of a self-powered drug release mechanism and theoretical predictions of the effects of uploaded drugs on polymer matrix structure using density functional theory and mechanical models in order to provide a more thorough understanding of the system. Table 1 Mechanical properties of PPy films.