Steric Effects in the Deposition Mode and Drug-Delivering Efficiency of Nanocapsule-Based Multilayer Films.

Using surface-initiated atom transfer radical polymerization (ATRP), block polymers with a series of quaternization degrees were coated on the surface of silica nanocapsules (SNCs) by the "grafting-from" technique. Molnupiravir, an antiviral medicine urgently approved for the treatment of SARS-CoV-2, was encapsulated in polymer-coated SNCs and further incorporated into well-defined films with polystyrene sulfonate (PSS) homopolymers by layer-by-layer (LBL) self-assembly via electrostatic interactions. We investigated the impact of the quaternization degree of the polymers and steric hindrance of functional groups on the growth mode, swelling/deswelling transition, and drug-delivering efficiency of the obtained LBL films. The SNCs were derived from coronas of parent block polymers of matched molecular weights-poly(N-isopropylacrylamide)-block-poly(N,N-dimethylaminoethyl methacrylate) (PNIPAM-b-PDMAEMA)-by quaternization with methyl sulfate. As revealed by the data results, SNCs with coronas with higher quaternization degrees resulted in a larger layering distance of the film structure because of weaker ionic pairing (due to the presence of a bulky methyl spacer) between SNCs and PSS. Interestingly, when comparing the drug release profile of the encapsulated drugs from SNC-based films, the release rate was slower in the case of capsule coronas with higher quaternization degrees because of the larger diffusion distance of the encapsulated drugs and stronger hydrophobic-hydrophobic interactions between SNCs and drug molecules.

Introduction

The ability to construct highly controlled layering structures enables important potential applications of layer-by-layer (LBL) films as antibacterial coatings and drug delivery systems.[1−7] LBL films containing strongly binding components, e.g., polystyrene sulfonate (PSS), are usually considered inherent nonequilibrium structures with irreversible bond polymer chains during LBL self-assembly.[8,9] These films could deliver a wide variety of functional reagents and controllably release active agents via external triggers.[10,11]

Functional composites based on inorganic nanoparticles have been recently developed for pharmaceutical applications.[12−14] Silica nanocapsules (SNCs) are one of the robust nanoparticles with potential applications in a wide range of fields due to their large surface area/volume ratio, low toxicity, and high cargo protection.[15,16] The buildup of nanofilms with SNCs as one of the constituents is an important approach for developing nanoscale delivering vehicles of therapeutic reagents. The development of core–shell nanocapsules typically involves the usage of organic solvents, which is potentially toxic to health. Hence, a mixture of dimethyl sulfoxide and coin oil was utilized as core materials to dissolve therapeutic reagents and prepare nanocapsules. The SNCs encapsulated with the antiviral drug were afterward adopted as one of the building blocks to develop nanoscale films with a defined layering structure by LBL deposition to release the drug in a controlled mode for targeted treatment.

Amphiphilic block copolymers are polymer chains with hydrophilic and hydrophobic blocks, which exhibit the advantage of good controllability, biocompatibility, and structural versatility.[17−19] They were able to possess convertible chain conformation, controlled interfacial properties, and a stimuli-responsive chemical structure. A combination of amphiphilic block copolymers with SNCs could efficiently enhance their stability, biocompatibility, and environmental sensitivity.[20,21] Nevertheless, the drug release rate could also be accelerated by functionalizing SNCs with block copolymers. Therefore, to obtain controlled drug delivery systems, it is critical to functionalize SNCs with amphiphilic block copolymers with regularly tailored chemical structures and physicochemical properties.[22,23] The LBL deposition allows for the buildup of the functional nanocomposites with environmental responsive properties for the controlled release of functional reagents from the defined-structure matrices.[24,25] However, few studies were conducted on the buildup of LBL films based on responsive block copolymer-functionalized SNCs for drug delivery.

Poly(N-isopropylacrylamide) (PNIPAM) is a biocompatible polymer with a lower critical solution temperature (LCST) at 32 °C. The PNIPAM-functionalized nanoparticles exhibit a temperature-triggered hydration–dehydration transition, which could be used for biomedical applications, such as temperature-modulated drug delivery systems. For instance, nanoparticles functionalized with poly(N-isopropylacrylamide)-block-poly(ethylene glycol) exhibited a temperature-triggered release of paclitaxel.[26] However, the nanoparticles functionalized with PNIPAM-based copolymers generally demonstrated a relatively fast release rate of drugs due to their high surface areas and irregular structures.[27] To overcome this issue, the poly(N-isopropylacrylamide)-block-poly(N,N-dimethylaminoethyl methacrylate) (PNIPAM-b-PDMAEMA) block copolymer was utilized to functionalize the SNCs and then embedded into thin films with a defined structure by LBL self-assembly to achieve the sustained drug release from the systems. Besides, the PNIPAM-b-PDMAEMA block copolymer was afterward quaternized to a certain quaternization degree to control the steric hindrance around the charged groups of PDMAEMA blocks in capsule coronas. The effect of the steric restriction on the interaction between the encapsulated drugs and film constituents was also evaluated to elucidate the molecular motion and relative release mode of the encapsulated drugs from the solid multilayer films.

Molecular mobility within multilayer films can sometimes be used to induce entrapment of therapeutic reagents. They also often allow potential applications of drug delivery systems that rely on molecular diffusion within the solid states for targeted treatment of diseases (such as coronavirus). For example, LBL films could be potentially used for multistage, multidrug delivery of therapeutic compounds. To achieve a sequential release of functional therapeutic compounds from LBL films, it is essential to incorporate strongly charged polyelectrolytes as barrier layers of LBL films to separate consecutive multilayer components and obtain film stratification.[28,29] However, the lack of understanding of molecular mobility inside the polyelectrolyte multilayers (PEMs) still limit precise control over the drug release profile. As with drug molecules in solution, the mobility of drug molecules within PEMs is affected by several factors, which modulate intermolecular binding.[30] For example, the transition of molecular motion from the non-Fickian diffusion to Fickian diffusion, commonly associated with enhanced intermolecular bondings, can be controlled by solution pH,[31] polymer molecular weight,[32] environmental temperature, and salt ions.[33] Among the variations, the charge density of the film constituents was considered one of the most important factors controlling drug release profiles and post-assembly entrapment.[34] For instance, the motion speed of the molecules within the solid films was increased as the ionization degree of polyamine species was reduced to a value below 70%.[29]

During the deposition process, heating the deposition solutions could lead to thicker LBL films and progressive molecular motions were observed after the films were immersed in the solutions.[35,36] Our group has recently reported faster diffusion of drugs out of the multilayer films with weak polyelectrolytes. Interestingly, a correlation between the film thickness and the intermolecular bonding was also observed.[37] Herein, we studied the dependence of growth behavior and the drug release profile of SNC-based multilayer films on the charge density, chain hydrophobicity, and steric restrictions of charged groups. While the impacts of these factors have been evaluated in homopolymer-based multilayer films,[38] the effects of these parameters on the internal structure of SNC-based LBL films and their drug release profile are still unexplored.

In this work, LBL films were constructed through electrostatic interactions using SNCs functionalized with block copolymers of systematically varied quaternization degrees and steric hindrance of functional groups. We demonstrate that intermolecular steric hindrance significantly affects not only the absolute value of the internal film thickness but also its temperature-triggered swelling/deswelling behaviors. Molnupiravir, an antiviral drug urgently approved by the U.S. Food and Drug Administration (FDA) to treat severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) disease, was efficiently encapsulated in the LBL films. Importantly, steric restrictions of the functional groups showed a remarkable impact on the mobility of drug molecules released from LBL films. The LBL films were able to slowly release the encapsulated molnupiravir during 80 days while retaining a constant morphology and thickness due to strong electrostatic interactions between polymer-functionalized SNCs and PSS chains. The incorporation of SNCs with quaternized coronas provides a tool for the fabrication and investigation of multilayered film formation, relevant for understanding molecular motions in a solid matrix as well as the basis for the controlled release of pharmaceutical reagents.

Experimental Section

Preparation of Molnupiravir-Encapsulated SNCs and LBL Films

Details of materials along with an exact description of the procedures for the development of SNCs and LBL films are given in the Supporting Information. The synthetic procedures of bare SNCs and polymer-coated SNCs encapsulated with molnupiravir are illustrated in Scheme .

Scheme 1

Schematic Representation for the Synthesis of Bare SNCs and SNC-g-PNIPAM-b-QPDMAEMA Loaded with Molnupiravir in the Cores

Characterization of the Polymer Coatings on Nanocapsules

The polymers functionalized on drug-encapsulated SNCs were measured by GPC and 1H-NMR techniques (Figures S1 and S2 in the Supporting Information).[39,40] PNIPAM-b-quaternized PDMAEMAs (PNIPAM-b-QPDMAEMAs) were determined to possess quaternization degrees of ∼20, ∼40, and ∼100%, which are abbreviated as PNIPAM-b-Q20M, PNIPAM-b-Q40M, and PNIPAM-b-Q100M, respectively.

Dynamic Light Scattering (DLS)

Hydrodynamic sizes, polydispersity indices, and ζ-potentials of bare and block polymer-coated SNCs were characterized in phosphate buffer solution (PBS) with a Zetasizer Nano-ZS using the Smoluchowski approximation.

Scanning Electron Microscopy (SEM)

The bare and block polymer-functionalized SNCs were dried on silica substrates and characterized by a Zeiss Auriga dual-beam FIB-SEM.

Nitrogen Adsorption–Desorption Isotherms (BET)

The N2 adsorption/desorption isotherms of SNCs were characterized by a Micromeritics ASAP 2020 system at 77 K. Five milliliters of SNC solutions at pH 5.5 were dispersed into 50 mL of methanol and stirred for 1 h at 25 °C. The samples were concentrated by centrifugation. The procedure was repeated three times to completely extract the loaded drugs and solvents. Afterward, the SNC solids were outgassed at 573 K for 4 h. The results are demonstrated in Figure S3.

Atomic Force Microscopy (AFM)

The morphology of the LBL films was characterized by the AFM technique using an NSCRIPTOR dip pen nanolithography research platform (NanoInk) and P-MAN-SICT-0 AFM cantilevers (Pacific Nanotechnology, Inc.) with a nominal 0.2 N/m force constant.

Swelling/Deswelling of LBL Films

The dry thickness of multilayer films was measured with a custom-built phase-modulated ellipsometer at an incidence angle of 70°. The refractive indices were input as 1.456 and 1.500 for silicon dioxide and dry films, respectively. The branched polyethylenimine (BPEI) precursor layer of the films was determined to possess a thickness of 1.6 ± 0.1 nm. To monitor the thickness of the wet films, the wafers were immersed in the 0.01 M PBS solution in a fluid quartz cell. The solution was tuned to the designed temperature, equilibrated for 20 min, and then measured by in situ ellipsometry.

Loading Capacity Measurement

To explore the contents of molnupiravir loaded in the LBL films, 1 cm × 1 cm silicon wafers coated with the films were placed in 10 mL of methanol solution to extract the encapsulated molnupiravir. The content of molnupiravir in the extraction was measured using a high-performance liquid chromatography (HPLC) system with an Inertsil YMC-ODS-AQ 3F C18 column, coupled to a Sciex API III+ triple quadrupole mass spectrometer. Standard solutions based on molnupiravir/DMSO/coin oil mixtures of known concentrations were adopted to create standard calibration curves. The drug-loading content (DLC) of the multilayer films was determined according to the formula

Temperature Effect on the Drug Release Profile

To explore the drug release rate of molnupiravir from free-standing SNCs, 1 mL of molnupiravir-loaded SNCs at a concentration of 0.5 mg/mL was filled into dialysis membrane bags (MWCO: 1000) and then dialyzed in 20 mL of 0.01 M pH 5.5 PBS at 25 and 37 °C, respectively. After certain time periods, the content of molnupiravir released in the PBS was determined by high-performance liquid chromatography/mass spectrometry (HPLC/MS).

To determine the drug release behavior of molnupiravir from the films, 1 cm × 1 cm silicon slides coated with multilayer films were immersed in 20 mL of 0.01 M pH 5.5 PBS in sealed glass vials. The temperature of the solutions was controlled at 25 and 37 °C. After a certain time interval, the contents of molnupiravir in the solutions were analyzed by HPLC/MS. The cumulative drug release was determined as W(t)/W(∞) × 100%, where W(t) is the cumulative weight of molnupiravir in the solution after a predetermined time period t and W(∞) is the total weight of molnupiravir entrapped in the films.

Statistical Analysis

All of the experiments were conducted in triplicate, and an unpaired t-test was adopted during statistical analysis. A p-value of 0.05 or less is considered significant.

Determination of Binding Energies between the Functional Groups in the Systems

To elucidate the effect of steric hindrance on the growth of LBL films, the calculation of the binding energy was simplified to the simulation of stable geometries of the functional groups of capsule coronas and PSS pairs. To take the water solvent into count, the PDMAEMA, Q100M, and PSS were treated as NH3+O(CH3)2((CH3)2CHCOOCH2CH2), N+H2O(CH3)3((CH3)2CHCOOCH2CH2), and SO3–C6H4CH2CH3, respectively. The stable geometries of the charged pairs of NH3+O(CH3)2((CH3)2CHCOOCH2CH2)/SO3–C6H4CH2CH3 and N+H2O(CH3)3((CH3)2CHCOOCH2CH2)/SO3–C6H4CH2CH3 were modeled using Gaussian 98 ab initio calculations, respectively.[41] In addition, to compare the binding energies between the diverse capsule coronas and the encapsulated drug, the theoretical simulation was simplified to calculate the stable geometries of the functional groups of the systems. The stable geometries of the functional group pairs of NH3+O(CH3)2((CH3)2CHCOOCH2CH2)/molnupiravir and N+H2O(CH3)3((CH3)2CHCOOCH2CH2)/molnupiravir were simulated using Gaussian 98 ab initio calculations, respectively. The simulations were conducted based on the Hartree–Fock SCF method, using the 6-31G split-valence basis set in the Gaussian orbital wave functions.

Results and Discussion

Drug-Loaded SNC Synthesis and Characterization

Nanocapsule-based composites are attractive to researchers because of their defined structure, controlled physicochemical properties, and high biocompatibility.[42,43] In our previous work, SNC/biopolymer composites were successfully developed via an interfacial sol–gel reaction and LBL self-assembly, which exhibited a slow release rate of the encapsulated drugs in response to environmental triggers.[44] Because it was difficult to control the molecular weight and environmental response of biopolymers, the amphiphilic block copolymer PNIPAM-b-PDMAEMA was chosen to coat the surface of SNCs by the “grafting-from” ATRP technique (Scheme ) so as to enhance their feasibility in hydrophobicity, surface charge, and response to external triggers. PNIPAM moieties of block copolymer coatings exhibit an LCST at ∼32 °C, and PDMAEMA moieties possess positively charged amino groups for charge compensation of the systems.[45] The nanoscale composites demonstrate both temperature response and positive charge. The hydrodynamic size, structural integrity, and molnupiravir loading efficiency of the capsules were enhanced by block copolymer coatings. Besides, the block copolymer coatings were partially quaternized to manipulate steric hindrance and hydrophobicity of the charged groups in SNC coronas to control the interaction between the capsules and drug molecules, as well as the drug release rate from the systems.

Table shows the hydrodynamic size and ζ-potential results of molnupiravir-loaded bare SNCs, PNIPAM-grafted SNCs (SNC-g-PNIPAM), PNIPAM-b-PDMAEMA-grafted-SNCs (SNC-g-PNIPAM-b-PDMAEMA), PNIPAM-b-Q20M-grafted SNCs (SNC-g-PNIPAM-b-Q20M), PNIPAM-b-Q40M-grafted SNCs (SNC-g-PNIPAM-b-Q40M), and PNIPAM-b-Q100M-grafted SNCs (SNC-g-PNIPAM-b-Q100M) at pH 5.5, respectively. The hydrodynamic size of bare SNCs was determined to be 162.3 ± 15.6 nm, which was in agreement with the size of oil/water emulsion droplets (168.2 ± 21.5 nm) during composite preparation. Surface functionalization of PNIPAM and PNIPAM-b-PDMAEMA lead to the enlargement in the hydrodynamic size of the capsules to 197.5 ± 20.3 and 251.2 ± 27.1 nm, respectively. The size difference indicated the growth of polymer chains on the surface of SNCs, which is beneficial to the drug-delivering efficiency of the therapeutic reagents. After the quaternization of PDMAEMA moieties in the capsule coronas, the hydrodynamic size of SNCs was slightly increased to 255–270 nm due to the repulsion force between the positively charged polymer brushes at the surface. The 1H-NMR spectra in Figure S2 confirmed the synthesis and quaternization of the polymer brushes on capsule surfaces. Besides, bare SNCs exhibited a weak surface charge with a ζ-potential of ∼+11 mV due to the amino groups at the surface introduced by γ-aminopropyl triethoxysilane (APTES) and cetyltrimethylammonium chloride (CTAC) components. The results of the BET measurement on the SNCs are presented in Figure S3. The SNCs showed pore diameters of 40–130 Å, indicating that the shell of SNCs was composed of mesoporous silica.[16] Moreover, SNCs demonstrated a large surface area (SBET = 371 m2/g and VBET = 0.93 cm3/g), which provides them with a high drug-loading capacity. The coating of neutrally charged PNIPAM had no significant effect on the surface charge, and the ζ-potential of SNCs was retained at ∼+13 mV. After the growth of PNIPAM-b-PDMAEMA at the surface of nanocapsules, the ζ-potential was significantly raised to ∼+35 mV due to the positively charged PDMAEMA moieties in capsule coronas. ζ-Potentials were elevated to around +38, +42, and +46 mV after the PDMAEMA moieties of the capsule coronas were quaternized to Q20M, Q40M, and Q100M, respectively, implying the successful quaternization of the polymers at the capsule surface. As shown in Figure , the diameter of the spherical nanocapsules in SEM images was increased along with the growth and quaternization of polymer chains, which is in agreement with the results obtained from DLS measurement. Interestingly, no agglomeration or precipitation of the nanocapsules was observed in the system, possibly due to strong repulsion between nanocapsules induced by highly charged CTAC surfactants and QPDMAEMA moieties at the surface.

Table 1

Average Values and Standard Deviations of the Hydrodynamic Size and ζ-Potential for 0.5 mg/mL Bare SNCs, SNC-g-PNIPAM, SNC-g-PNIPAM-b-PDMAEMA, SNC-g-PNIPAM-b-Q20M, SNC-g-PNIPAM-b-Q40M, and SNC-g-PNIPAM-b-Q100M in 0.01 M PBS with a pH of 5.5 at 25 °C

sample	hydrodynamic size (nm)	ζ-potential (mV)
bare SNCs	162.3 ± 15.6	11.38 ± 2.51
SNC-g-PNIPAM	197.5 ± 20.3	13.19 ± 2.73
SNC-g-PNIPAM-b-PDMAEMA	251.2 ± 27.1	35.63 ± 5.13
SNC-g-PNIPAM-b-Q20M	257.3 ± 28.2	38.26 ± 5.85
SNC-g-PNIPAM-b-Q40M	263.2 ± 29.3	42.57 ± 6.21
SNC-g-PNIPAM-b-Q100M	270.5 ± 31.2	46.21 ± 6.58

Figure 1

SEM images of drug-loaded (a) bare SNCs, (b) SNC-g-PNIPAM, (c) SNC-g-PNIPAM-b-PDMAEMA, (d) SNC-g-PNIPAM-b-Q20M, (e) SNC-g-PNIPAM-b-Q40M, and (f) SNC-g-PNIPAM-b-Q100M air-dried on the surface of oxidized silicon wafers at pH 5.5, respectively.

Figure shows SEM images of bare SNCs, SNC-g-PNIPAM, SNC-g-PNIPAM-b-PDMAEMA, SNC-g-PNIPAM-b-Q20M, SNC-g-PNIPAM-b-Q40M, and SNC-g-PNIPAM-b-Q100M at the surface of silicon wafers. The average lateral size of PNIPAM-coated SNCs was determined to be 180–220 nm, which was larger than that of bare SNCs (150–180 nm) due to polymer brushes grafted on the surface of SNCs by covalent bonding. The integrity of spherical capsules in a vacuum environment reflected the good drying stability of SNCs with polymer coatings. In addition, the growth and further quaternization of PDMAEMA moieties in the coronas of capsules led to the enlargement of average lateral sizes of the capsules on substrates to 250–320 nm in Figure c–f. The hydrodynamic sizes of PNIPAM-b-quaternized PDMAEMA-grafted-SNCs (SNCs-g-PNIPAM-b-QPDMAEMA) with a series of quaternization degrees (0, 20, 40, and 100%) are shown in Figure S4, respectively. When the environmental temperature was lower than the LCST of PNIPAM (32 °C), the hydrodynamic sizes of SNCs-g-PNIPAM-b-QPDMAEMA ranged from 240 to 270 nm, respectively. Interestingly, the hydrodynamic sizes were increased along with the enhancement of the quaternization of the PDMAEMA moieties, possibly due to strong electrostatic repulsions in capsule coronas. When the temperature was elevated above 32 °C, the SNCs-g-PNIPAM-b-Q100M composites underwent temperature-induced deswelling of the capsule coronas, leading to the reduction in the average hydrodynamic sizes (210–240 nm), respectively. PNIPAM moieties became dehydrated when the temperature was higher than its LCST, resulting in a reduced hydrodynamic size for the composites. The PNIPAM moieties also demonstrated reversible hydration/dehydration behaviors so that the hydrodynamic size of the capsules went up to 240–270 nm with the reduction of the temperature to a value below its LCST, respectively. The reversibility of such swelling/deswelling behaviors of the composites was assured by the strong covalent binding of the polymer brushes to SNC shells. Based on SEM images in Figure , the lateral sizes of nanocomposites at the surface were larger than their hydrodynamic sizes in solution in Figure S4. The difference was possibly due to the collapse of the capsules at the surface after air-drying steps, which was also observed for nanoparticle/polymer composites in other reports.[46,47] Due to the strong repulsion between the charged coronas of the composites, no aggregation or agglomeration was seen during the DLS measurement.

Multilayer Buildup of SNC/PSS Films

SNCs-g-PNIPAM-b-QPDMAEMA nanocomposites and PSS polymers were utilized as building constituents to obtain thin films via LBL self-assembly. To improve the adhesion of multilayer films to the substrate surface, a BEPI precursor layer was deposited on silicon wafers to enhance the attachment of the films to the wafer surface. As illustrated in Scheme , the multilayer films were deposited on the substrate by LBL self-assembly based on strong electrostatic interactions between positively charged polymer-functionalized SNCs and negatively charged PSS polymers. pKa values are 6.6 and 2.3 for PDMAEMA and PSS, respectively, so that the pH values of all of the solutions were controlled at 5.5 throughout the deposition process.[48,49] The environmental conditions ensured strong electrostatic interactions between the highly charged capsule and PSS components, which are critical for the structural integrity of the films during the assembling process.[50]

Scheme 2

Schematic Representation of LBL Deposition of SNC/PSS Films Based on the Electrostatic Interaction between Polymer-Coated SNCs and PSS Homopolymers

The thickness of SNC/PSS multilayer films was measured by ellipsometry during the deposition process. Figure demonstrates the thickness of LBL films as a function of the bilayer number. All of the multilayer films were assembled in a linear growth trend, indicating strong bonding between the film constituents. According to the charge compensation mechanism, a certain part of building constituents should be adsorbed on the oppositely charged layer so as to achieve a constant surface charge density for the systems. The attachment of the constant number of building blocks to the previously adsorbed layer assured the robust assembly of LBL films without the presence of competitive displacement or component intermixing, leading to a defined internal layering structure of the films.[51,52]

Figure 2

Ellipsometric thickness of dry LBL films composed of SNC-g-PNIPAM-b-PDMAEMA/PSS (squares), SNC-g-PNIPAM-b-Q20M/PSS (circles), SNC-g-PNIPAM-b-Q40M/PSS (triangles), and SNC-g-PNIPAM-b-Q100M/PSS (pentagons) on silicon substrates as a function of the bilayer number.

Based on the growth of the LBL films in Figure , the slopes of the curves were calculated to determine the thickness of each layer in the films. The thicknesses of the bilayer in SNC-g-PNIPAM-b-PDMAEMA/PSS, SNC-g-PNIPAM-b-Q20M/PSS, SNC-g-PNIPAM-b-Q40M/PSS, and SNC-g-PNIPAM-b-Q100M/PSS films were calculated to be ∼37, ∼40, ∼51, and ∼63 nm, respectively. The average thickness of the PSS layer was ∼4 nm in all of the films by ellipsometry. The dry thickness of the SNC monolayer in all of the films was 3–4 times smaller than the hydrodynamic size of SNCs, probably due to the loose packing conformation of SNCs at the surface and dehydration behaviors of polymer coronas after the drying process.[53,54] The layer thickness was increased as a function of the quaternization degree of the QPDMAEMA moieties in the films. This significant difference resulted from the steric bulk of the methyl groups at the quaternary nitrogen of QDPMAEMA. The stable geometries for the charged group pairs of NH3+O(CH3)2((CH3)2CHCOOCH2CH2)/SO3–C6H4CH2CH3 and N+H2O(CH3)3((CH3)2CHCOOCH2CH2)/SO3–C6H4CH2CH3 calculated by Gaussian 98 using Hartree–Fock SCF modeling and the 6-31G split-valence basis set are shown in Figure S5 in the Supporting Information. The binding energies for the two pairs are summarized in Table S1. The binding energy in NH3+O(CH3)2((CH3)2CHCOOCH2CH2)/SO3–C6H4CH2CH3 pairs was larger than that in N+H2O(CH3)3((CH3)2CHCOOCH2CH2)/SO3–C6H4CH2CH3 pairs. It implied that the distance between oppositely charged groups was enhanced by the steric bulk around the amino group of QPDMAEMA moieties. Such an enhancement in steric hindrance led to reduced binding energies between oppositely charged layers and enlarged the intermolecular distance between the amino group in QPDMAEMA chains and the sulfonate group in PSS chains in the LBL films with higher quaternization degrees. The results were also observed in the previous reports on the QPDMAEMA-based LBL films.[38] The enhancement of the film thickness was found in the LBL films deposited with homopolymer pairs with high steric hindrance. The results of neutron reflectometry and AFM analyses pointed out that the multilayers possessed the elevated thickness after the building block PDMAEMA was quaternized.[55,56] Incomplete surface coverage of capsules in one-bilayer films was also observed in AFM images in Figure S6a–d. According to AFM topography images, the RMS roughness of the films could be calculated. SNC-g-PNIPAM-b-PDMAEMA/PSS, SNC-g-PNIPAM-b-Q20M/PSS, SNC-g-PNIPAM-b-Q40M/PSS, and SNC-g-PNIPAM-b-Q100M/PSS one-bilayer films possessed the RMS roughness of 27.3 ± 2.9, 30.7 ± 3.8, 32.1 ± 4.0, and 36.5 ± 4.6 nm, respectively. The difference between the RMS roughness and layer thickness probably resulted from the strong electrostatic repulsion between charged SNCs and the absence of mobility of adsorbed capsules at the surface, which is critical for the formation of the closely packed particle structure. The results from the observation were previously reported in other nanoparticle-based LBL films.[57] The attachment of the later-arriving capsules was prohibited by the insufficient spaces for touching spots between previously adsorbed capsules. Besides, the lateral size of nanocapsules at the surface in AFM images was also larger than their hydrodynamic size in solution, which resulted from the strong interaction between capsule coronas and the PSS layer and the collapse of the polymer coronas of the nanocomposites after air-drying.[51]

Compared to one-bilayer films in Figure S6, AFM topography images of SNC-g-PNIPAM-b-QPDMAEMA/PSS 3-bilayer films demonstrated a closely packed spherical morphology in Figure . From the analysis of AFM topography, the RMS roughnesses of 3-bilayer films of SNC-g-PNIPAM-b-QPDMAEMA/PSS with the quaternization degrees of 0, 20, 40, and 100% were determined as 35.1 ± 4.5, 39.3 ± 4.7, 45.5 ± 6.1, and 53.5 ± 7.2 nm, respectively. The RMS roughnesses of 3-bilayer films were larger than the corresponding one-bilayer films in Figure S6, indicating that the spaces between loosely packed capsules in previously adsorbed layers were filled up by the later-arriving capsules and PSS layers. Besides, the RMS thickness of 3-bilayer films was parity with their average bilayer thickness, indicating the adsorption of certain amounts of the particles at a quasi-3D surface with a constant roughness per deposition cycle and the resulting linear growth mode for the LBL films. In addition, RMS roughnesses of the LBL films with the same bilayer numbers were increased along with the enhancement of the quaternization degree of PDMAEMA moieties in capsule coronas, respectively. The phenomenon was also reported in other studies related to the internal structure of capsule-based films.[51] It was possibly induced by the enlargement of the hydrodynamic size of SNC-g-PNIPAM-b-QPDMAEMA and the reduction in the electrostatic interaction between capsule coronas and PSS layers. The quaternization of amino groups of PDMAEMA moieties led to stronger steric hindrance and a weaker electrostatic interaction between SNC coronas and PSS chains, which resulted in the enhanced layer thickness for SNC/PSS films.

Figure 3

AFM topography images of typical (a) [SNC-g-PNIPAM-b-PDMAEMA/PSS]3, (b) [SNC-g-PNIPAM-b-Q20M/PSS]3, (c) [SNC-g-PNIPAM-b-Q40M/PSS]3, and (d) [SNC-g-PNIPAM-b-Q100M/PSS]3 films deposited from 0.2 mg/mL solutions controlled at pH 5.5. Scale: 2 μm × 2 μm.

Temperature-Manipulated Swelling–Deswelling Behaviors of [SNC/PSS]3 Films

In situ phase-modulated ellipsometry was adopted to explore the swelling/deswelling patterns of [SNC-g-PNIPAM-b-PDMAEMA/PSS]3, [SNC-g-PNIPAM-b-Q20M/PSS]3, [SNC-g-PNIPAM-b-Q40M/PSS]3, and [SNC-g-PNIPAM-b-Q100M/PSS]3 films as a function of environmental temperature, respectively. The ellipsometric analysis of the films was carried out by decoupling their thicknesses and refractive indices.[58] Silicon substrates coated with LBL films were immersed in 0.01 M PBS in quartz cells and equilibrated for 10 min before the ellipsometric measurement. All of the solutions were controlled at pH 5.5.

The effect of environmental temperature on the thicknesses and the corresponding refractive indices of wet LBL films with various quaternization degrees are shown in Figures and S7. The in situ AFM technique was also utilized to characterize the surface morphology of the wet films. At 37 °C, the average thicknesses of the [SNC-g-PNIPAM-b-PDMAEMA/PSS]3, [SNC-g-PNIPAM-b-Q20M/PSS]3, [SNC-g-PNIPAM-b-Q40M/PSS]3, and [SNC-g-PNIPAM-b-Q100M/PSS]3 films were 125 ± 16, 135 ± 19, 170 ± 25, and 205 ± 28 nm and the corresponding refractive indices were 1.424, 1.416, 1.410, and 1.405. Compared to dry film thicknesses, the thicknesses of wet films were slightly larger, indicating the adsorption of a small amount of water in the matrix and gentle swelling of the films at 37 °C (swelling ratios of 1.27, 1.20, 1.17, and 1.08, respectively). While the solution temperature was reduced down to 25 °C, LBL films demonstrated remarkable swelling behaviors due to the hydration of PNIPAM moieties in capsule coronas. The thicknesses of the corresponding wet films were determined to be 158 ± 21, 170 ± 23, 208 ± 28, and 252 ± 33 nm (swelling ratios of 1.61, 1.51, 1.43, and 1.33) at 25 °C and the refractive indices were reduced to 1.381, 1.368, 1.352, and 1.349, respectively. The transition in the film thickness and refractive index in temperature-manipulating cycles was reversible, indicating their robust structural integrity after several repeated cycles. The reversible swelling/deswelling behaviors of PNIPAM moieties in the films ensured long-term temperature control over their structure and morphology.

Figure 4

Reversible temperature-triggered swelling/deswelling of [SNC-g-PNIPAM-b-PDMAEMA/PSS]3 (squares), [SNC-g-PNIPAM-b-Q20M/PSS]3 (circles), [SNC-g-PNIPAM-b-Q40M/PSS]3 (triangles), and [SNC-g-PNIPAM-b-Q100M/PSS]3 (pentagons) films in 0.01 M PBS at 25 and 37 °C, respectively, measured by in situ ellipsometry. All of the solutions were controlled at pH 5.5.

The surface morphology of the wet LBL films was analyzed by in situ AFM after three temperature-tuning cycles. The AFM topography images in Figure S8a–h show that all of the four films demonstrated a constant spherical structure after three cycles. Based on AFM images of wet [SNC-g-PNIPAM-b-PDMAEMA/PSS]3 films in Figure S8a, the average lateral size of capsules and RMS roughness of films were calculated to be 200–250 nm and 29.3 ± 4.6 nm at 37 °C, respectively. While the environmental temperature was reduced to 25 °C, the enlargement of the capsule’s lateral size (300–350 nm) and the reduction of RMS roughness (15.1 ± 2.9 nm) were observed in AFM topography images in Figure S8b. AFM images of [SNC-g-PNIPAM-b-Q20M/PSS]3, [SNC-g-PNIPAM-b-Q40M/PSS]3, and [SNC-g-PNIPAM-b-Q100M/PSS]3 films in Figure S8c–h demonstrated similar temperature-triggered transition in the lateral capsule size and RMS roughness of the films. When the temperature was below its LCST, PNIPAM moieties in capsule coronas in the films became hydrophilic and hydrated, resulting in the significant swelling of the capsules and smooth surface morphology. Note that LBL self-assembly with a top PSS layer as the coating is an effective approach to stabilize the SNCs against desorption. Under temperature triggers, PNIPAM-based constituents might be detached from LBL films.[59] The un-cross-linked LBL films herein maintained a constant morphology and thickness in various environmental conditions, which were advantageous over the PNIPAM-incorporated LBL films in previous reports. Robust LBL films with reversible temperature-responsive behaviors were prepared by a simple LBL deposition method, which was ensured by strong covalent bonding between polymer brushes and the SNC surface and electrostatic interactions between oppositely charged SNC coronas and PSS chains.

Effect of Environmental Temperature on the Delivery Efficacy of Molnupiravir

The drug-loading capacity of the films and the release behaviors of molnupiravir from the films were also detected to determine their drug-delivering efficacy. Molnupiravir is an antiviral medication that was developed at Emory University and later owned by Merck & Co. It is an isopropylester prodrug of the synthetic nucleoside derivative N4-hydroxycytidine, which gets hydrolyzed to an intermediate EIDD-1931 and allocated to tissues, where it is transformed into an active 5′-triphosphate by a host kinase. It works against viruses by introducing copying errors during viral RNA replication, a process known as viral error catastrophe. Molnupiravir has been urgently approved for the treatment of SARS-CoV-2.[60]

Polymer-coated SNCs were prepared with the molnupiravir/DMSO/coin oil mixture in the core and then embedded into LBL films so that molnupiravir was successfully loaded into the films. To determine their loading capacity, the films were treated with methanol to completely extract the loaded molnupiravir. The drug content in methanol solution was further analyzed by the HPLC/MS technique according to calibration curves of molnupiravir/methanol standards with known concentrations. The results demonstrated that 8.95 mg/m2 molnupiravir was loaded in the [SNC-g-PNIPAM-b-PDMAEMA/PSS]3 films. Additionally, [SNC-g-PNIPAM-b-Q20M/PSS]3, [SNC-g-PNIPAM-b-Q40M/PSS]3, and [SNC-g-PNIPAM-b-Q100M/PSS]3 LBL films were able to encapsulate 11.56, 12.31, and 12.63 mg/m2 of molnupiravir, respectively. The difference in the drug-loading capacity of the LBL films resulted from enhanced surface coverage of the capsules in LBL films with higher quaternization degrees. Based on the capsule content in the LBL films, the loading capacity results reflect (3.5–4.0) × 104 molnupiravir molecules per capsule for all of the LBL films, implying that >90% of overall molnupiravir was encapsulated in the capsules in all of the films. The SNC-embedded LBL films showed a remarkably higher drug-loading capacity than nanoparticle-based LBL films in other studies.[61]

The release profiles of molnupiravir from 3-bilayer LBL films were monitored by the HPLC/MS technique. The silica wafers coated with the films were incubated in PBS at 25 and 37 °C, respectively, and the content of molnupiravir in PBS was measured at a certain time interval. All of the solutions were maintained at pH 5.5 throughout the measurement to ensure the structural integrity of the films. Figure a,b demonstrates the release profiles of molnupiravir from [SNC-g-PNIPAM-b-PDMAEMA/PSS]3, [SNC-g-PNIPAM-b-Q20M/PSS]3, [SNC-g-PNIPAM-b-Q20M/PSS]3, and [SNC-g-PNIPAM-b-Q100M/PSS]3 films at 25 and 37 °C, respectively. At 37 °C, it could be observed that around 81, 76, 62, and 45% of molnupiravir was released from [SNC-g-PNIPAM-b-PDMAEMA/PSS]3, [SNC-g-PNIPAM-b-Q20M/PSS]3, [SNC-g-PNIPAM-b-Q40M/PSS]3, and [SNC-g-PNIPAM-b-Q100M/PSS]3 films after 80 days, respectively. All of the SNC-embedded films demonstrated a long-term release mode of molnupiravir and the release rates were slower than those from homopolymer-based LBL films in previous reports.[62] The attachment of PSS chains to SNCs in the LBL films enhanced the retention of molnupiravir in capsule cores and reduced the mobility of the drug molecules in the LBL films. A large amount of molnupiravir was entrapped in all 3-bilayer films after 80 days. The phenomenon is possibly induced by hydrophobic–hydrophobic interactions between the drug molecules and the film constituents, as well as the slow motion of hydrophobic molnupiravir molecules in the film matrix. As the temperature was controlled above the LCST of PNIPAM, the diffusion of hydrophobic molnupiravir molecules was reduced by enhanced hydrophobic–hydrophobic interactions between dehydrated PNIPAM moieties in the films and molnupiravir. Besides, the drug release was accelerated when the temperature was reduced to a value below the LCST of PNIPAM. ∼91, ∼89, ∼87, and ∼82% of molnupiravir were released from [SNC-g-PNIPAM-b-PDMAEMA/PSS]3, [SNC-g-PNIPAM-b-Q20M/PSS]3, [SNC-g-PNIPAM-b-Q40M/PSS]3, and [SNC-g-PNIPAM-b-Q100M/PSS]3 films in 20 days at 25 °C, respectively. The acceleration of the release rate of molnupiravir was possibly due to temperature-induced hydration of PNIPAM moieties in LBL films. Note that certain amounts of molnupiravir were irreversibly entrapped in the LBL films for all of the samples, which was probably due to the strong bonding of drug molecules to the hydrophobic cores of nanocapsules and polymer complex-based hydrophobic domains.[59] At the temperature below the LCST of PNIPAM, capsule coronas became hydrated and adsorbed a large amount of water, which resulted in high pressure on silica capsules, a loose polymer network in LBL films, and weaker bonding to molnupiravir. All of the changes based on the polymer swelling led to faster diffusion of molnupiravir out of capsule cores and LBL films. Besides, irreversible entrapment of a small amount of molnupiravir in hydrophobic domains of LBL films at 25 °C was also implied by the incomplete release of the entrapped molnupiravir from the LBL films. Note that the release rate of molnupiravir from LBL films was accelerated by increasing the quaternization degree of PDMAEMA moieties. The phenomenon was also observed in QPDMAEMA homopolymer-based LBL films.[38] Molecular diffusion in the direction parallel and vertical to the substrates could be slowed down along with the increase of the quaternization degree of the QPDMAEMA moieties in the LBL films. Moreover, the exchange and displacement of functional agents embedded within either hydrogen-bonded or electrostatically interacted LBL films was affected by the environmental pH and temperature conditions.[63,64] In our case, enhanced steric hindrance in the amino group of QPDMAEMA moieties along with the increase of quaternization degree led to larger capsule coronas, a thicker film layer, and a longer molecular diffusion distance for [SNC-g-PNIPAM-b-QPDMAEMA/PSS]3 films. Hence, the release rate of molnupiravir out of QPDMAEMA-embedded films was reduced as a function of the quaternization degree of PDMAEMA moieties.

Figure 5

Release kinetics of molnupiravir from [SNC-g-PNIPAM-b-PDMAEMA/PSS]3 (squares), [SNC-g-PNIPAM-b-Q20M/PSS]3 (circles), [SNC-g-PNIPAM-b-Q40M/PSS]3 (triangles), and [SNC-g-PNIPAM-b-Q100M/PSS]3 (pentagons) films in 0.01 M pH 5.5 PBS controlled at (a) 25 °C and (b) 37 °C, respectively. (c) Schematic representation of temperature-induced swelling/deswelling behaviors of the films and the related drug release modes.

The release behaviors of molnupiravir from free-standing nanocapsules into solutions are demonstrated in Figure S9. At 37 °C, ∼91, ∼88, ∼85, and ∼78% of the molnupiravir was released out of the cores of free-standing SNC-g-PNIPAM-b-PDMAEMA, SNC-g-PNIPAM-b-Q20M, SNC-g-PNIPAM-b-Q40M, and SNC-g-PNIPAM-b-Q100M within 20 days, respectively. When the temperature value was raised above its LCST, the collapse and dehydration of PNIPAM on the surface of SNCs led to strong hydrophobic–hydrophobic bondings between molnupiravir and capsule coronas, resulting in slow diffusion of the drug molecules out of the nanocapsules. When the temperature was reduced below the LCST of PNIPAM (25 °C), hydrophobic–hydrophobic interaction between PNIPAM moieties and drug molecules became weaker due to reduced hydrophobicity of PNIPAM. In addition, swelling of PNIPAM moieties at 25 °C resulted in a loose corona network of nanocapsules and a faster diffusion rate of molnupiravir out of the capsule cores. Therefore, the release rate of molnupiravir out of free-standing nanocapsules at 25 °C was dramatically higher than that of molnupiravir at 37 °C, with ∼93, ∼91, ∼91, and ∼89% molnupiravir released within 10 days for the corresponding free-standing naocapsules. Compared to that from free-standing nanocapsules, the release rate of molnupiravir from the LBL films built up with nanocapsules was significantly slower since the motion of molnupiravir was restricted by the boundaries between SNC and PSS layers in LBL films. AFM topography images of the LBL films after 80-day incubation are demonstrated in Figure S10. The [SNC-g-PNIPAM-b-PDMAEMA/PSS]3, [SNC-g-PNIPAM-b-Q20M/PSS]3, [SNC-g-PNIPAM-b-Q40M/PSS]3, and [SNC-g-PNIPAM-b-Q100M/PSS]3 films retained their spherical morphology upon exposure to PBS for 80 days due to strong electrostatic interactions between capsule coronas and PSS layers within LBL films.

The release rate of molnupiravir from free-standing SNCs and SNC-embedded films was decelerated along with the enhancement of the quaternization degree of PDMAEMA moieties. The phenomena indicated that the steric restriction caused by the quaternization of PDMAEMA had a significant impact on the entrapment of the drugs. The steric bulk of the methyl groups at the quaternary nitrogen of QPDMAEMA led to the diverse bonding of the hydrophobic molnupiravir to components of the free-standing SNCs and LBL films, typically QPDMAEMA moieties. The stable geometries for the group pairs of NH3+O(CH3)2((CH3)2CHCOOCH2CH2)/molnupiravir and N+H2O(CH3)3((CH3)2CHCOOCH2CH2)/molnupiravir were simulated via Gaussian 98 based on Hartree–Fock SCF modeling and the 6-31G split-valence basis set. The geometric results based on the simulation are shown in Figure S11, and the binding energies between the two groups are summarized in Table S2 in the Supporting Information, respectively. The binding energy in the NH3+O(CH3)2((CH3)2CHCOOCH2CH2)/molnupiravir pair was smaller than that of the N+H2O(CH3)3((CH3)2CHCOOCH2CH2)/molnupiravir pair. The difference in the binding energies supports the hypothesis that the steric bulk around the amino group enhanced the hydrophobicity of QPDMAEMA chains and the resulting hydrophobic–hydrophobic interactions between QPDMAEMA moieties and molnupiravir.

The mechanism of the drug release mode from the LBL films was also studied according to the plots of log W(t)/W(∞) against log t in Figure S12, where W(t) is the weight of molnupiravir released after a time period t, W(∞) is the total weight of molnupiravir entrapped in the systems, and W(t)/W(∞) is the ratio of released molnupiravir. The slopes of the plots are listed in Table S3 in the Supporting Information. The slopes of the plots for the [SNC-g-PNIPAM-b-PDMAEMA/PSS]3 films were determined to be 0.455 and 0.639 at 37 and 25 °C, respectively. These results indicated that the release mode of molnupiravir from the films was dependent on the environmental temperature. At temperatures above the LCST of the PNIPAM block (37 °C), the molnupiravir was released from the system based on Fickian diffusion (n < 0.5).[38] On the other hand, non-Fickian diffusion was observed for the release of molnupiravir (0.5 < n < 1) at 25 °C since the hydration of PNIPAM reduced the bonding of the drug to the hydrophobic domains of the LBL films.[65] The transition in drug diffusion mode triggered by temperature variation also occurred in [SNC-g-PNIPAM-b-QPDMAEMA/PSS]3 films. The reason for the incomplete release of molnupiravir from the LBL films was because of permanent entrapment of the hydrophobic drug in capsule cores and hydrophobic domains induced by the interaction between the capsule and PSS layers in LBL films. The phenomenon agrees with the results in previous reports. In addition, it was found by other groups that the variation of the pore size in the capsule wall and matrix tunnels in LBL films resulted in diverse release profiles of the entrapped drugs from the systems.[66,67] In our case, a long-term release profile of entrapped therapeutic reagents was observed for the LBL films built up with polymer-functionalized SNCs and PSS homopolymers. The impact of temperature on the release profile of the drug molecules from LBL films resulted from the swelling/deswelling transition of PNIPAM moieties in the LBL films triggered by temperature variation. Remarkable acceleration of the molnupiravir release rate from the LBL films was observed when the temperature was reduced below the LCST of PNIPAM moieties. The controllability and sustainability of the structure and drug release profile of SNC/PSS LBL films make them a promising platform for in vitro sustained delivery systems of therapeutic reagents.

Conclusions

Based on the ATRP “grafting-from” technique, nanocapsule composites with controlled hydrodynamic size, surface functionalization, and environmental response were developed to encapsulate molnupiravir. The nanocapsule composites and PSS homopolymers were further utilized as building blocks to develop electrostatically interacted multilayer films by LBL self-assembly. The growth modes and swelling behaviors of the films were dependent on the quaternization degree of the positively charged polymer moieties and the corresponding steric hindrance of functional groups in capsule coronas. The SNC/PSS films exhibited a high loading capacity of molnupiravir and temperature-triggered swelling/deswelling behaviors, leading to a long-term release profile of the entrapped drug in response to temperature stimulus. The electrostatically interacted LBL films could be potentially used for in vitro controlled/sustained release of therapeutic reagents approved for the treatment of SARS-CoV-2. These nanofilms can be combined with nanocomposites with the upper critical solution temperature (UCST) to build up potential platforms for in vivo drug delivery systems.