Scalable microfabrication of drug-loaded core–shell tablets from a single erodible polymer with adjustable release profiles

1-Hydroxycyclohexyl phenyl ketone (photo-initiator, 0.1 wt%) was weighed and placed into a 15 ml Falcon tube. Then, PNA was transferred into the tube, followed by the mixture of EGDT and PETMP. The solution was mixed to obtain a homogenous mixture. The initial mole ratio of PNA to the total amount of monomers containing thiol groups (EGDT and PETMP) was 100:100. In this study, four different initial mole ratios of PNA to PETMP and EGDT were prepared and used: PNA: PETMP: EGDT = 100:100:0, 100:75:25, 100:50:50, and 100:25:75. The homogenous pre-polymer solution was then purged under the inert gas (nitrogen or argon) for 3 min and was transferred to PDMS molds. PDMS molds with different designs and geometries were fabricated and used according to the desired geometry of each polymer. The pre-polymer solution in the molds was exposed to ultraviolet (UV) light (CL-1000 UVP Cross-linker) equipped with 365 nm UV lamps (intensity = $\sim$ 5 mW cm⁻²) for 5 min. To study the release profiles of dye-loaded tablets, a model compound (orange G) was added to the pre-polymer solution of some of the tablets. Orange G (1–3 wt%) was weighed and added to the purged solution and mixed using sonication for 25 min. Similar to the pre-polymer without the model compound, the orange G dye-loaded solution was transferred to PDMS molds and synthesized under the same UV light for 5–15 min. The reaction scheme is presented in the Supplementary file (supporting figure S1 (available online at stacks.iop.org/BF/12/045007/mmedia)).

3D printed master molds were used to fabricate the PDMS molds on which the pre-polymer solutions were polymerized. The master molds were printed by a commercial digital light processing (DLP) 3D printer (PICO2, ASIGA). Desired 3D models were designed by AutoCAD software and were printed layer by layer from a UV-curable resin (Clear 2500T). The layer thickness was set to 250 µm, and the UV-curing time for each layer was 5 s. The 3D printed objects were treated using sonication in ethanol for 10 min and washed by deionized (DI) water to remove any uncured polymer.

The tablet has two main parts: the core and the shell (figure 1(A)). The core is made of a surface eroding polymer loaded with a drug (shown in orange in figures 1(A) and (B) while the same polymer without any drug is used for the shell (the grey color in figure 1). The core is enclosed by the shell and is placed closer to the top surface compared to the other faces. When the thickness of the shell is 'large enough' for all sides except the top, the release of the drug is ensured only from the top surface of the tablet (figure 1(B)). This design relies on the surface eroding behavior of the polymer which preserves the tablet structure and integrity during erosion and ensures the erosion of the polymer at a similar rate from all sides. An experiment was conducted that verified the latter (see section 2.4). Therefore, in this design, the water first reaches the drug at the top. Because the thickness of the shell is large enough, the drug is released completely from the top before the water reaches it from the other sides. Accordingly, the release profile of the drug is only governed by the variations in the surface area of the drug-loaded core polymer along with its height (figure 1(C)). As an example, in figure 1, the core polymer's surface area from the top is increasing along its height, showing that this tablet is designed for obtaining an increasing release profile (figure 1(D)). A similar approach was used to design the constant (figure 7(A)) and decreasing-increasing (figure 8(A)) release profile tablets.

Zoom In Zoom Out Reset image size

The 600 µm distance from the upper edge of the tablet core to the top surface is considered to prevent the initial burst release of the drug upon contact with biological fluids. The surface eroding PAs used in this work showed an induction period in which they absorb water by diffusion before surface erosion starts. The erosion front (the thickness that water diffuses to the thiol-ene PA prior to the start of the surface erosion process) is around ∼400 µm. The 600 µm distance also ensures that both the drug release and the surface erosion occur concurrently.

To ensure that the polymer erodes at similar rates from all sides, the erosion of cylindrical thiol-ene PAs was studied. Four thiol-ene PA cylinders with initial mole ratios of 100:100:0, 100:75:25, 100:50:50, and 100:25:75 were synthesized in cylindrical PDMS molds. The PDMS molds were fabricated using 3D printed master molds with the diameter and height both equal to 8.7 mm. The synthesized polymers (cylindrical tablets) were immersed in 10 ml PBS (pH = 7.4) in glass vials, before being placed on a shaker (VWR micro-plate shaker). The shaker was set up at room temperature and the shaking rate of 120 rpm. The diameter and height of the tablets were measured frequently (approximately one every hour) during the erosion. For each measurement, tablets were removed from the solution and their surfaces were slowly wiped using disposable paper wipers. The dried tablets were then put on a printed coloured graph paper, and photos were taken using a Canon SX201 IS digital camera from four different sides of each tablet, before putting the tablet back in PBS. Images from the top and bottom of the tablet were used to monitor the diameter reduction, while the two images from the sides provided information on variations in the tablet's height. The NIH ImageJ software (version 1.8.0) was used to measure tablet dimensions from the photos taken. Experiments were repeated three times for each polymer tablet.

A microfluidic network that was fabricated using the conventional standard photo- and soft-lithography techniques [30, 31] was utilized to create the dye-loaded polymers used as the core of the tablets. The SU-8 photoresist was poured on a four-inch silicon wafer, and a 300 μm-thick layer of photoresist was spin-coated (Laurell Tech Corp.) on the top surface of the silicon substrate. Considering the photo- and soft-lithography procedure for microfabrication, a photomask that replicated the final desired shape is used. In the design for the core of the tablet, instead of using the shape of a single tablet core for the photomask, a throughput pattern of tablet cores was designed using AutoCAD software which was then printed as the photomask on a high-transparent sheet (supporting figure S2(A)). The SU-8 photoresist (with the desired thickness spin-coated on the substrate) was covered by the photomask and exposed to high-intensity UV-light (AB-M Inc.) for 80 s. When using a photo-mask, the UV-light only passes through the UV-transparent parts, crosslinking only the desired areas on SU-8. Therefore, the SU-8 photoresist polymer hardened only at the designed pattern and the other parts remained uncross-linked. Lastly, SU-8 developer was used to remove the uncross-linked photoresist from the wafer, yielding the final SU-8 mold (supporting figure S2(B)).

Figure 2(A) shows an example of the design for scalable manufacturing of the tablet core for an increasing release profile. First, the SU-8 master molds were fabricated using the above-mentioned steps (figure 2((A)i–ii)). To fabricate the microfluidic device, the PDMS polymer mixture containing the 1:10 ratio of base to curing agent was thoroughly mixed. To remove the trapped air bubbles created in the PDMS sample, the mixture was degassed in a vacuum chamber for 15 min. As shown in figure 2((A)iii), the negative PDMS mold is developed and peeled off from the SU-8 positive mold. To facilitate the peel-off process, the SU-8 mold was silanized to make the mold's micrometer-sized channels more hydrophobic. The positive SU-8 mold was silanized using (tridecafluoro-1,1,2,2- tetrahydrooctyl) trichlorosilane by putting the mold in a sealed petri-dish on a hot plate for 2 hrs at 65 °C. To make the PDMS replica mold, the mixture was poured on top of the SU-8 mold and were cured in an oven for 2 hrs at 80 °C. Finally, the cured PDMS was easily peeled off from the SU-8 master mold, yielding the throughput patterns connected through the microfluidic network (figure 2((A)iii)).

Zoom In Zoom Out Reset image size

A platform for using the fabricated PDMS negative molds to make the polymeric tablet cores was designed (figure 2((A)iv)). Another PDMS negative mold with a pattern mirrored to that of the original was fabricated and placed on top of the original PDMS negative mold, making a microfluidic network with 600 µm in thickness. The top PDMS negative mold was punched on two sides with a 2 mm biopsy punch (EMS-Core Sampling Tools) to create inlet/outlet ports through which solutions flow into the channels created by the two PDMS negative molds (figure 2((A)iv)). The two PDMS layers were sandwiched between the two rigid PMMA sheets (3.3 mm thickness). Uniformly distributed holes were designed using AutoCAD software and were cut on PMMA sheets using the CO₂ laser (SUNCOO K40 laser cutter). These holes were designed for using screw-nuts through which a uniform force was applied to squeeze the two PDMS layers together to prevent any leakage. At the center of the top PMMA sheet, a pattern was cut using the laser cutter to allow exposure of the solution in the PDMS layers to UV-light (figure 2((A)iv)) for curing the tablet cores.

Pre-polymer solutions were prepared as described in the polymer preparation procedure section. In this study, two pre-polymer solutions with initial mole ratios of PNA: PETMP: EGDT = 100:100:0, and 100:75:25 were prepared. Approximately 5 ml of solutions containing 1 wt% of orange G model compound were transferred into a vial. The solution was continuously injected into the microfluidic network from the inlet port using the Tygon Microbore tubing (0.079''ID) (figure 2((B)i)). A vacuum pump was connected to the outlet, helping draw the solution to fill the chambers. After the solution filled the chambers completely, the features were exposed to UV-light (UVP Crosslinker, intensity = $\sim$ 5 mW cm⁻²) for 5 min. The screws were then opened and the three connected dye-loaded tablet cores were easily peeled off by separating the two PDMS layers.

The cured micro-fabricated features (the three connected tablet cores) were then placed in PDMS wells (figure 2((B)ii)) fabricated using master molds printed by commercial DLP 3D printer. The procedures for the fabrication of the 3D printed master molds were the same as the steps described in section 2.2. The 3D model of the wells was designed so that when the throughput micro-fabricated cores are placed in the wells, each of the cores stays at the center (from the sides, horizontally) and at a pre-determined distance from the bottom and the top of the well (vertical alignment). Particularly, as explained in the tablet design section, the cores were closest to the top and far enough from all other sides of the wells. The dimensions of wells (cylindrical molds with diameter × height equal to 8.7 × 6.6 mm, 8.7 × 8.7 mm, and 14.1 × 12 mm) were determined based on dimensions of the tablet cores to provide desired thickness as tablet shells. After placing the orange G dye-loaded cores in the wells, the empty spaces around the polymers inside the wells were filled using the same pre-polymer solution as the core but without the orange G dye. Another 5 min UV-exposure was conducted to cure the newly added polymer solution. Finally, three tablets were peeled off from the PDMS wells (figure 2((B)ii)) and simply were separated using a blade by cutting the connecting part. To continue reusing the PDMS molds, the channels were washed after peeling off the dye-loaded polymer using the mixture of water and ethanol. After washing the channels carefully, they were dried using high-pressure air for 2 min which ensures cleaning the molds from dirt and possible residues.

The designs of tablet cores, shown in supporting figure S3, were printed as photomasks and the SU-8 master molds were fabricated through photo-lithography steps (figure 2((C)i)). The PDMS molds were successfully fabricated via a standard soft-lithography process using the SU-8 master molds (figure 2((C)i)). The depth of the chambers on PDMS negative molds was measured using a microscope (Nikon 334 Eclipse Ti-E). The measured depths ranged from 298 µm to 301 µm for different molds that are very close to the designed value (300 µm), showing successful fabrication of PDMS negative molds.

The original PDMS layer and the one with the mirrored pattern were put on top of each other to create 600 µm cavities between them (figure 2((C)ii)). Typically, two PDMS layers irreversibly bound to each other or a PDMS layer bonds to a glass slide using standard oxygen plasma activation process to make a microfluidic network [32–34]. The pattern can be destroyed by the formation of an irreversible seal between the PDMS mold and a glass slide [35]. However, in this study, there was a desire to access the features and remove the orange G dye-loaded PA solution once it is cured in the network. Therefore, to seal the channels and at the same time prevent the leakage of the fluid, at first, small clamps were used to squeeze the two layers together as shown in figure 2((C)iii). However, leakage of the polymer solution from the microfluidic channels was observed using the clamps. Alternatively, two rigid PMMA sheets were utilized to sandwich the PDMS layers. Screws and nuts were used to apply uniform force and to secure the alignment. This platform was designed and fabricated for creating the tablet cores. The successful injection of the pre-polymer solution containing the orange G model compound via inlet/outlet ports is shown in figure 2((C)iv).

The ultimate goal of this study is to achieve different release profiles through a new tablet design fabricated by a throughput fabrication platform. To meet this goal, different tablet geometries were designed using AutoCAD software. Different designs of the throughput tablet cores which are connected to each other are shown in supporting figure S3. These tablet cores were designed for increasing (supporting figure S3(A)), and constant release profiles (supporting figure S3(B)). The width of the tablet cores is increasing and uniform from the top along their heights in the increasing and constant tablets, respectively. Supporting figure S3(C) shows different designs for the tablet core with constant release profiles that have two, three, or four constant release cores in each connected feature (this design is described in details in the next section). These designs were printed as photomasks and went through soft-lithography steps for making the PDMS molds containing connected features as a microfluidic network.

The micro-fabricated cores for the increasing and constant release profiles were placed into PDMS wells where the pre-polymer solution without drug was poured and polymerized under the UV-light. The three tablets for each geometry were placed in 5 ml PBS at pH = 7.4 separately before being placed on a shaker with a shaking rate of 120 rpm and at 37 °C. Almost every hour, 750 µl of each PBS solution was withdrawn, transferred into three wells of a 96-well plate (each well contained 250 µl), and analyzed using a microplate reader (Asys UVM 340). The light intensity of the orange G model compound in a 250 µl sample was measured using the absorbance measured by a the microplate reader. The concentrations were then calculated using the concentration-absorption calibration curve of orange G in the wavelength with the maximum absorbance of UV light (475 nm) (Supporting figure S4). For the UV microplate reader device, the absorbance-concentration relation was calculated using $A = \alpha \,C$ where A is the absorbance of the UV light by the model compound and C is the concentration of the model compound in the sample. The absorbance is linearly related to the concentration. The slope (α in $A = \alpha \,C$) was calculated for the micro-plate reader to be 27.13 by fitting a linear line to absorbance-concentration data.

The absorbance of each sample was measured using the UV plate reader, and the concentration of the model compound in the sample was calculated using $A = \alpha \,C$. An equivalent amount of the fresh PBS (750 µl) was added after withdrawing the same amount of solution. After collecting the concentration data of the model compound at certain time intervals, the fractional release (equation (1)) and the release rates (equation (2)) of the model compound from each tablet were calculated. In equation (1), M₀ is initial amount of the orange G dye dispersed in the tablet (which is known), C_i is concentration of the model compound calculated for the i^th sample at time t, V_t is total volume of the sample (e.g. 5 ml in this study), and V_i is withdrawn sample volume (250 µl per well or 750 µl per sample), and F_i% is fractional released percentage of the dye.

$$\begin{equation}{F_i}\,\% \, = \,\frac{{{C_i} \times {V_t} + \,\mathop \sum \nolimits_1^{i - 1} {C_j}{V_i}}}{{{M_0}}} \times 100.\end{equation} \tag{ 1 }$$

In equation (2), R_i is release rate of the model compound at time t for the i^th sample, and T_i is time at which the i^th sample was withdrawn.

$$\begin{equation}{R_i} = \frac{{\left( {{F_i} - {F_{i - 1}}} \right) \times {M_0}}}{{{T_i} - {T_{i - 1}}}}.\end{equation} \tag{ 2 }$$

The tablet designs were modified to improve the capacity of the tablet in containing higher quantities of therapeutics. One-arm, two-arm, and four-arm designs for the core of the tablets were devised to make use of extra spaces in the shell of cylindrical tablets. In the modified designs, instead of using only the axial direction for releasing the dye, the radial directions were used as well by loading the dye in two-arm and four-arm tablet designs. Figures 3(A)–(C) shows the schematic front view illustrations of the one-, two-, and four-arm tablet designs, respectively. Although in the one-arm design, the core is closest to the top of the tablet (same as the tablet designs described so far), in the two and four-arm designs, the core is as close to the right, left, and bottom of the tablet as to the top side. It means the water could access the orange G dye from one, two, and four sides for the tablets shown in figures 3(A), (B) and (C), respectively.

Zoom In Zoom Out Reset image size

As was shown in supporting figure S3(C), similar to the increasing and constant designs, the modified tablets are designed in a throughput connected pattern. The tablet cores were printed as photomasks and fabricated using the throughput microfabrication method previously discussed. Similar to what was described above, the tablets' shells were fabricated with the help of PDMS cylindrical wells. Three separated tablets for each design were immersed in PBS solution for the release test using the same procedure mentioned above. The concentration of the model compound was calculated using the absorbance values measured by the microplate reader. The fractional percentage model compound released and the release rates for these three modified designs were calculated using equations (1) and (2) and then compared to each other.

Several studies have shown that PAs predominantly undergo surface erosion [36, 37]. It has been shown that the linear mass loss profile of PAs results in a near zero-order drug release from these polymers [9, 38]. Most of these studies used PAs in the form of a slab (a thin layer of polymer). Using the new type of PAs synthesized by thiol-ene photopolymerization, a rectangular slab (2 × 10 × 10 mm) made of PNA and PETMP showed a linear mass loss profile. More complicated mass loss profiles were observed for a cube made of the same polymer [39]. The non-linearity observed in the mass loss profile of the more complex thiol-ene PA geometries (such as cubes and cylinder) does not contradict the surface-eroding behavior of the polymer. Surface erosion behavior is defined as the quality to preserve the tablet geometry during the erosion time. This unchanging geometry of tablets is attributed to a similar dimension reduction rates in all directions. In our study, the diameter and height variations for each tablet were monitored and calculated from the photos that were taken at specific time intervals (approximately once every hour).

The average of the measured diameters and heights from triplicate experiments for each tablet were plotted over time (figure 4) Where the linear behavior of the dimension reduction for cylindrical tablets is shown. An increased initial mole ratio of EGDT to PETMP in the thiol-ene PAs network results in higher erosion rates. The rates of the reduction for both diameter (figure 4(C)) and height (figure 4(D)) of all the tablets were calculated by finding the slope of the linear curves fitted to the measured data. Table 1 summarizes the results of the reduction rates for diameters and heights of these four tablets. The dimension reduction profiles in figure 4 showed that the reduction of diameters and heights for each tablet is linear. In addition, similar diameter and height reduction rates are observed for tablets made of the same PA. Having the same reduction rate in different dimensions is a very crucial characteristic for a polymer because it increases its applications in controlled-release DDSs. The adjustable release rates from the tablet designs rely on similar reduction rates in different dimensions.

Zoom In Zoom Out Reset image size

The linear behavior of the reduction rates (the calculated correlation coefficients R² ≈ 1) observed in both dimensions (shown in figures 4(A) and (B)), allowed calculating the expected dimension reduction rate of the same polymer for a new cylindrical design. Ideally, this expected reduction rate can be calculated by measuring the dimensions of the tablet at least at two arbitrary time points during erosion and fitting a linear function to dimension versus time data. The slope of the dimension versus time represents the reduction rate. This linear behavior of the reduction rate in this study eliminates the need for monitoring tablet dimensions during the erosion time. For example, by having the initial and final dimensions and the erosion time, the reduction rates can be calculated and used in the tablet design for releasing pre-determined patterns of drugs. The diameter and height reduction rates for the PAs used to fabricate tablets in this study at physiological temperature (37 °C) were monitored and calculated using the same approach. The reduction rates for the heights and diameters of the thiol-ene PA cylindrical tablet with initial mole ratio of PNA:PETMP:EGDT = 100:100:0 were calculated to be similar and equal to 0.34 mm hr⁻¹ and 0.66 mm hr⁻¹ for the PA cylindrical tablet with the initial mole ratio of PNA:PETMP:EGDT = 100:75:25.

To create the tablet shell, PDMS wells were successfully fabricated using 3D printed master molds. Firstly, cubic wells were designed in a way that the distance between every two adjacent cubes was the same as the distance between the two adjacent micro-fabricated tablet cores. The orange G dye-loaded tablet cores were placed at the center of the wells at a certain height using external aligners. Two aligners were used to hold the fabricated dye-containing part at the specific x,y,z position. Figure 5(A) shows the printed cubic wells, the PDMS negative molds, and the three cubic surface eroding tablets created in the cubic PDMS wells (from left to right). Figure 5(B) shows an alternative design for the cubic wells in AutoCAD in which two aligners were added at both ends of the 3D printed object. Two aligners were designed with certain heights and a notch on them where the micro-fabricated tablet cores hang to ensure the right x,y,z position. These two aligners facilitate the correct placement of the fabricated dye-containing part in the right location inside the well, without the need for external aligners. All the tablet shell designs were changed to cylindrical objects (figure 5(C)) since tablets with cubic shapes have sharp corners that are inconvenient to swallow when used as oral tablets. The optimized design was the cylindrical wells with two aligners in both sides and two more aligners between every two adjacent wells (figure 5(D)). Aligners between wells were added to prevent the micro-fabricated part from bending or collapsing due to gravity.

Zoom In Zoom Out Reset image size

The resolution of the method used for fabrication of the tablet core is of higher importance than that of the tablet shell. The geometries of the micron-sized tablet cores determine the release profile of the orange G model compound and are also more complex than the geometries of the milimeter-sized PDMS wells that form the tablet shell. However, our tablet design benefits from a high-resolution technique for producing PDMS wells (DLP 3D printing). This 3D printing method is also capable of printing various geometries with sharp corners or curvatures, including cubes, cylinders, and spheres as the common form of tablets in the pharmaceutical industry. The higher surface quality of the tablet shells also ensures more precise control over the erosion rate of the tablet in the novel design proposed in this study. Lithography-based 3D printers, including stereolithography (SLA) and DLP 3D printers can create objects with more complex designs at a higher speed and higher resolutions compared to the other commercially available 3D printers such as fused deposition modeling (FDM) machines [40]. In this study, a DLP 3D printer (PICO2, ASIGA) was used to fabricate the master molds because of its fast and high-resolution manufacturing process compared to the SLA counterparts. The limitation of the lithography-based 3D printers is the limited number of the available UV-curable resins. Clear 2500T resin was used after evaluating its thermal stability in an oven for 2 hrs at 80 °C, which is the same thermal condition for curing the PDMS on top of the printed master molds.

The cores and shells of the tablets were fabricated using the developed platform described above for the throughput manufacturing of the tablets. As an example, the results of the throughput fabrication of the increasing release profile tablets are shown in figure 6(A), where it shows the final device developed for fabricating the orange G dye-loaded cores of the tablets. Three tablet cores for the increasing release profile that were fabricated and peeled off from the μm-sized PDMS mold are also shown in figure 6. The cross-linked dye-loaded features were set in place using the notch aligners embedded in the PDMS wells. These features were created using the 3D printed master molds (figure 6(B)). Empty spaces in the wells were filled by the same surface eroding polymer without the orange G model compound. All three wells were then exposed to the UV-light, and the final connected tablets were peeled off from the PDMS wells (figure 6(C)). These connected tablets were simply separated to obtain three individual tablets for the increasing release profile (figure 6(D)).

Zoom In Zoom Out Reset image size

Two of the PA compositions (PETMP:EGDT = 100:0, and 75:25) were used in fabrication of the oral tablets with adjustable release profiles. The other two polymers (PETMP:EGDT = 50:50, and 25:75) indicated fast erosion rates in comparision to the polymers were used (∼10 h vs 24 h). The 24 h window allows to release therapeutics in about one day in a controllable manner. The proposed tablet design in this study was tested at first for the constant release profile. The constant release profile tablets were designed such that the surface area of the tablet core from the top along its height was constant (figure 7(A)). Figure 7(B) shows the constant release profile core–shell tablet made of PA (initial mole ratio PNA:PETMP:EGDT = 100:100:0) in PBS solution after the first half of the core is eroded. The concentration of the orange G model compound in the sample was measured approximately once every hour and the fractional cumulative release of the model compound was calculated and plotted over time (figure 7(C)). The results show the linear fractional release of the model compound. On the top left on the graph in figure 7(C), the release rate of the orange G model compound is plotted. The release rate remain unchanged over time which agrees with the pre-determined constant release profile intended by the tablet design.

Zoom In Zoom Out Reset image size

To demonstrate the versatility of the tablet design, a different core–shell tablet was designed and fabricated (figures 8(A) and (B)). The core was fabricated for a decreasing and then increasing release profile of the orange G model compound. The surface area of the tablet core from the top along its height is decreasing and then increasing. Figure 8(C) shows the fabricated core–shell tablet for decreasing-increasing release profile (figure 8(B)) made of PA (initial mole ratio of PNA:PETMP:EGDT = 100:75:25) in PBS solution after the first half of the erosion. The fractional cumulative release of the model compound was calculated and plotted (figure 8(D)) over time. The fractional release of the model compound shows two distinct sections, one during the erosion of the decreasing part and the other happens during the erosion of the increasing part. The data for these two parts were fitted to two quadratic models using MATLAB MathWorks software. On the top left on the graph in figure 8(D), the release rates of the orange G model compound were plotted which shows a decreasing and then an increasing release rate over time.

Zoom In Zoom Out Reset image size

Release rates calculated for the constant and the decreasing-increasing release profile core–shell tablets showed the feasibility of tuning the pattern of release of the orange G model compound from the polymer using the proposed core–shell tablet design. The release profile patterns were only dependent on the tablet's core geometry, particularly the variations in its surface area from the top along its height as well as the location of the tablet core inside the shell. The linear reduction rates of the erodible tablets' dimensions make it possible to estimate the release profiles of the drugs from the tablets. Compared to other studies conducted on obtaining desired release profiles from polymeric tablets [41–43], the presented method is more straightforward and does not require complex modeling and/or solving of complicated equations for designing tablets with adjustable release profiles.

The drug-loading capacity of a polymeric delivery system is the total amount of drug that can be loaded in the system and depends on the polymer, drug, fabrication method, and the design of the DDS. In the tablet design, a higher intensity UV-light source was required to cross-link the polymers containing higher amounts of the orange G model compound (or any other therapeutics). The high-intensity UV-light transfers very high energy which may damage the loaded therapeutics. By modifying the tablet design and utilizing the extra space in the shell polymer, higher amounts of the model compound can be loaded into the tablets without increasing the total tablet volumes.

The height and diameter reduction rates for polymers used in this study were measured before and results showed the similarity of the reduction rates. One-arm, two-arm, and four-arm core–shell tablets were fabricated (figure 9(A)) using the throughput platform. The same weight percentage of the orange G was added to these three tablet designs so that the total initial amount of the orange G dye loaded in the two-arm and four-arm tablets were two and four times higher than the one-arm loaded dye, respectively. The cumulative release of the orange G from all three tablets was calculated and the mean values from triplicate experiments were plotted as shown in figure 9(B). Linear curves were fitted to data and slopes were calculated, showing the constant release rates as expected (considering the tablet core designs). Figures 9(C)–(E) showed the tablets fabricated for constant release profiles while they are eroding in PBS.

Zoom In Zoom Out Reset image size

Cumulative release of the orange G over time was linear for all three fabricated tablets which provided evidence for a constant release profile. The slopes of the linear lines were calculated to show the constant release rates (figure 9(B)). The release rate of the four-arm tablet was almost four times higher than the release rate of the one-arm tablet, and the slope of the linear cumulative release of the two-arm tablet showed an almost two fold increase compared to the one-arm tablet. These findings also imply that using the four-arm design, it was possible to release the same amount of the drug by loading almost four-times lower weight percentage of the drug compared to the one-arm tablet. Consequently, the intensity of the UV light source required for curing the photocrosslinkable polymeric tablet cores is reduced which lowers the risk of damages to loaded drugs. Moreover, the results of this experiment showed that higher amounts of drug could be released using the other sides of the tablet as well as the top side, while preserving the desired release profile.

Previous studies have shown an increasing loading capacity of controlled release DDSs by changing the polymer compositions [44]. Various fabrication techniques such as hot-melt extrusion and injection molding were also used to increase the loading capacity of the polymeric tablets [45–47]. In this study, using the modified tablet designs which utilize more dimensions of the tablet for releasing the therapeutics, the loading capacity of the designed tablets was increased by approximately four times, in a straightforward fashion.