Mesoporous Matrices as a Promising New Generation of Carriers for Multipolymorphic Active Pharmaceutical Ingredient Aripiprazole

The enhancement of the properties (i.e., poor solubility and low bioavailability) of currently available active pharmaceutical ingredients (APIs) is one of the major goals of modern pharmaceutical sciences. Among different strategies, a novel and innovative route to reach this milestone seems to be the application of nanotechnology, especially the incorporation of APIs into porous membranes composed of pores of nanometric size and made of nontoxic materials. Therefore, in this work, taking the antipsychotic API aripiprazole (APZ) infiltrated into various types of mesoporous matrices (anodic aluminum oxide, native, and silanized silica) characterized by similar pore diameters (d = 8–10 nm) as an example, we showed the advantage of incorporated systems in comparison to the bulk substance considering the crystallization kinetics, molecular dynamics, and physical stability. Calorimetric investigations supported by the temperature-dependent X-ray diffraction measurements revealed that in the bulk system the recrystallization of polymorph III, which next is converted to the mixture of forms IV and I, is visible, while in the case of confined samples polymorphic forms I and III of APZ are produced upon heating of the molten API with different rates. Importantly, the two-step crystallization observed in thermograms obtained for the API infiltrated into native silica templates may suggest crystal formation by the interfacial and core molecules. Furthermore, dielectric studies enabled us to conclude that there is no trace of crystallization of spatially restricted API during one month of storage at T = 298 K. Finally, we found that in contrast to the crystalline and amorphous bulk samples, all examined confined systems show a logarithmic increase in API dissolution over time (very close to a prolonged release effect) without any sign of precipitation. Our data demonstrated that mesoporous matrices appear to be interesting candidates as carriers for unstable amorphous APIs, like APZ. In addition to protecting them against crystallization, they can provide the desired prolonged release effect, which may increase the drug concentration in the blood (resulting in higher bioavailability). We believe that the “nanostructirization” in terms of the application of porous membranes as a novel generation of drug carriers might open unique perspectives in the further development of drugs characterized by prolonged release.


Detailed procedures of the confined samples' preparation
Firstly, all types of membranes were dried in an oven at T = 423 K under vacuum (10 -2 bar) for t = 24 h to remove any volatile impurities from the nanochannels before filling.Next, the cooled and weighed unfilled membranes were top-coated with the crystalline APZ.
Subsequently, the whole system was kept at T = 420 K for at least 4-5 h, until the mass of membranes increased to constant values.During this time, the melted API flowed inside the nanochannels by infiltration forces.After completing this process, the surface of all membranes was heated again up to T = 420 K at 0.1 MPa to remove the excess sample on the surface with a metal blade and a paper dust-free wipe.
Table S1.Values of T g T c , T m1 , and T m2 determined from non-isothermal DSC measurements for bulk APZ.S3.Values of both T g (T g,core , T g,interfacial ),T c , T m1 , and T m2 determined from nonisothermal DSC measurements for APZ within silanized silica of d = 8 nm.S4.Values of T g (T g,core , T g,interfacial ),T c , T m1 , and T m2 determined from non-isothermal DSC measurements for APZ within AAO of d = 10 nm.In the case of secondary relaxation processes, the Arrhenius power law was used to fit the temperature dependences of and (solid red lines in Fig. S5), and consequently,     determine the activation barrier ( where = β, γ): where is the pre-exponential factor, while R is the gas constant.

𝜏 ∞
Moreover, in order to determine the molecular origin of both secondary processes in APZ, especially check, which one is the true Johari-Goldstein (JG) relaxation of intermolecular origin, we applied the Coupling Model (CM) by Ngai. 1,2This approach assumes the following relation between the JG relaxation time ( ) and the primitive relaxation time ( ) of the CM: The value of can be determined from the parameters and the stretched exponent of the  0

KWW function ( =
, where is a coupling parameter) by the formula: is a constant, which is equal to 2 ps for most polymeric and low-molecular-weight glass   formers. 1,3We calculated (open stars in Fig. S5) at several temperatures close to T g from the  0 corresponding (T) of APZ and found that they are close to experimental β-relaxation times.

𝜏 𝛼
Hence, the good correspondence between and indicates that the -mode is a true JG- 0   process, whose source are local, non-cooperative motions of the entire API molecules.In turn, the faster γ-relaxation is a non-JG process of intramolecular character, which probably originates from the rotations of polar groups of APZ molecule.

Fig. S1 .
Fig. S1.Kissinger plot for exothermic crystallization peaks observed for APZ incorporated into native silica templates.

Fig. S2 .
Fig. S2.Evolution of XRD patterns on heating of APZ in AAO pores.

Fig. S5 .
Fig. S5.Relaxation map of bulk APZ.The open stars are the primitive relaxation times (τ 0 ) of the CM calculated with n=0.44 (β KWW =0.56) at several temperatures close to T g from the corresponding experimental τ α and values given by the VFT fit.

Table S2 .
Values of both T g (T g,core , T g,interfacial ), T c , T m1 , and T m2 determined from nonisothermal DSC measurements for APZ within native silica of d = 8 nm.