Morphology, topography and mechanical properties of the MH-loaded fibrous

MH-loaded, aqueous viscous PVA solutions of two different compositions were prepared, and fibrous samples were manufactured with electrospinning. After the optimization of the fiber formation process parameters and compositions of the viscous

solutions for the electrospinning process, well-defined, round-shaped samples were achieved on the collector. Randomly oriented, clear fibrous structure without any beads and film-like areas can be observed in the SEM images of each sample (Figure 16).

Figure 16 SEM photos of the poly(vinyl alcohol)-based, metoclopramide-HCl-loaded electrospun sample either containing polysorbate 80 (PVA-PS-MH) (A, B) or

hydroxypropyl-β-cyclodextrin (PVA-CD-MH) (C, D) (Magnification: 3500× (A, C) and 10000× (B, D)

On the SEM images, clusters or any visible signs of heterogeneity could not be detected in either sample e.g., surface crystallization of the active pharmaceutical ingredient and other artifacts was missing. In the case of the PVA-PS-MH sample, a more densely packed fibrous structure was achieved, while a sparse coverage was obtained for PVA-CD-MH fibers.

Figure 17 Histograms of the fiber diameter distribution of metoclopramide-HCl-loaded electrospun sample either containing polysorbate 80 (PVA-PS-MH) (A) or

hydroxypropyl-β-cyclodextrin (PVA-CD-MH) (B)

336±88 nm and 323±62 nm (mean ±SD) average fiber diameter values were achieved for PVA-PS-MH and PVA-CD-MH electrospun samples, respectively. In the case of the PVA-PS-MH sample, a slightly wider fiber distribution was obtained than that of the PVA-CD-MH fibers with a more uniform structure (Figure 17).

Fiber diameter distributions of the two electrospun samples were investigated using the Kolmogorov–Smirnov test, which indicated that distributions of PVA-PS-MH and PVA-CD-MH fibers were normal (p = 0.844 and 0.416, respectively). A one-way analysis of variance (ANOVA) revealed that no significant difference between the diameter of the two formulations was found (p=0.211).

Figure 18 AFM amplitude-contrast images of poly(vinyl alcohol)-based, metoclopramide-HCl-loaded, either containing polysorbate 80 (PVA-PS-MH) (A, B, C)

or hydroxypropyl-β-cyclodextrin (PVA-CD-MH)(D, E, F) nanofibers.

Panel B and E are higher resolution images of areas indicated by yellow rectangles on panel A and D, respectively. Height sections (C, F) taken alongside the green dashed

line in panel B and E, respectively.

AFM topography of both PVA-PS-MH (Figure 18A and B) and PVA-CD-MH (Figure 18D and E) fibrous sample appeared similar to those seen in SEM images (Figure 16). The PVA-CD-MH fibers appeared as long, curved, continuous, cylinders.

In the case of PVA-PS-MH fibers, some ribbon-like, flat structure can also be observed (see top left and bottom right corners in Figure 18A).

The height of the investigated regions of the fibers varied between 330-360 nm and 300-450 nm for PVA-PS-MH and PVA-CD-MH fibers, respectively. Relatively smooth fiber surface with only a few nm height variations was observed in the case of both MH-loaded nanofibrous formulations. The height profile of PVA-PS-MH fibers (Figure 18C) showed sharper peaks than the cyclodextrin containing ones, which had a more rounded profile plot (Figure 18F), indicating that PVA-PS-MH fibers display a bit coarser surface.

For the investigation of the mechanical properties of the PVA-based nanofibers formulated with different excipients by AFM, 100 nN, 500 nN, 1 µN and 5 µN loads were

applied at well-defined points of the fiber surface, that was pressed and retracted with the AFM tip at a constant 1 µm/s velocity.

Figure 19 Effect of the mechanomanipulation: AFM-amplitude contrast images of metoclopramide-HCl-loaded electrospun fibers either containing polysorbate 80 (PVA-PS-MH) (A) or hydroxypropyl-β-cyclodextrin (PVA-CD-MH) (B) taken after force spectroscopy. The yellow rectangles indicate areas where force maps (10 × 10 force curves) were taken from. The maximum load applied in each region is written next to

the rectangles.

The manipulated area was re-scanned. No noticeable alternations were found in regions loaded with 100 nN forces (Figure 19). At 500 nN load, few small depressions were seen for PVA-CD-MH fibers, while in the case of the PVA-PS-MH sample, no effect was detected.

Nevertheless, in both formulations at 1 µN and 5 µN loads, permanent surface depressions were observed; thus, these forces enabled plastic deformation. Shallower and blurred depressions were obtained in PS-MH fibers (Figure 19A), while in PVA-CD-MH fibers, those were deeper and had a more definite profile reflecting the shape of AFM tips (Figure 19B).

Figure 20 shows the representative force curves that were taken at different maximum loads.

Figure 20 Representative force curves taken from metoclopramide-HCl-loaded either polysorbate 80 containing (left panels) or hydroxypropyl-β-cyclodextrin containing (right panels) nanofibers at 100 nN (A, B), 500 nN (C, D), 1 µN (E, F), 5 µN (G, H) maximum loads. Horizontal axes show the distance from the fiber surface.

The negative peak in approach curves in case of 100 nN maximum load can refer to a considerable attractive interaction between the fibers and the tip.

At 0 nm distance, where the fiber surface was reached, the force followed an apparent linear increase up to the threshold load (~100 nN), which is a sign of the elastic response.

When the tip was retracted, the force was decreased, and the retraction curve was roughly parallel to the approach curve. Only a minimal hysteresis can be observed between the two curves, suggesting that at this load there was no plastic deformation.

The large negative force peak of the retraction curve is the consequence of the tip-fiber adhesion, and then the force reached 0 as the tip was drawn farther from the fiber surface.

The ascending force region of approach curves was fitted with the Hertz-model of elasticity (96) that adapted to AFM force spectroscopy (Figure 20A and B dotted lines).

According to distributions of Young's modulus calculated from the fits, in both cases apparent normal distribution was achieved, and for PVA-PS-MH and PVA-CD-MH fibers, the mean±SD values were 3.26±1.74 GPa and 1.48±0.90 GPa, respectively (Figure 21A).

Figure 21 Histogram of Young-moduli and plasticity of the metoclopramide-HCl-loaded fibers formed either polysorbate 80 containing (PVA-PS-MH) or

hydroxypropyl-β-cyclodextrin (PVA-CD-MH)

With the increasing loads, the hysteresis area between the approach and the retraction curves was larger, which was referred to as a plastic deformation of the fibers. In the case of the PVA-CD-MH sample, a larger hysteresis area can be observed at each applied load;

thus, this fibrous formulation is more plastic (plasticity distributions at 5 µN load can see in Figure 21B, as it was concluded from Figure 19.