3.1. Stereolithographic Printing Process
The dosage forms in this study were fabricated utilizing the AM technique of SLA. SLA is a method in which objects may be fabricated or produced by using focused ultraviolet (UV) light and photocurable materials to produce solid 3D forms due to a photopolymerization reaction occurring. Photopolymerization is the use of light in order to initiate a polymerization reaction, thus resulting in the conversion of a liquid monomer to a solid polymer through a chemical reaction taking place [28
]. The polymerization reaction is initiated by the addition of light-sensitive compounds called photoinitiators, which become active under the appropriate wavelengths, creating free radicals due to the conversion of the absorbed light energy [29
]. These free radicals are then consumed in the reaction converting the liquid monomer into a solid-state, which may be referred to as a crosslinked hydrogel [28
]. The photoinitiator used within this study is Irgacure TPO, a Norrish Type 1, photoinitiator which is a single-molecule and undergoes photocleavage into radical fragments when exposed to light from an appropriate wavelength as illustrated in Figure 2
In order to fabricate the drug dosage forms, it was necessary to prepare the photopolymer solutions to be loaded into the SLA 3D printer which was carried out by altering the ratio of PCL Triol to PEGDA in the formulation with the ratios being chosen based on previous experimentation work in our lab and on the rheological profiling of existing formulations.
PCL Triol is a biodegradable, semi-crystalline, aliphatic polyester, with low molecular weight and melting point which has been investigated to a great extent as a biomaterial as well as in drug delivery applications [32
]. One reason for this is the degradation products of PCL being naturally occurring metabolites in the human body with sutures composed of PCL being approved by the US Food and Drug Administration (FDA) [35
]. Further to this, low molecular weight PCL Triols have been investigated for their potential to act as plasticizers [33
] with Kanis et al. [34
] indicating that it is the three hydroxyl groups present which allow it to act as a plasticizer. Furthermore, PCL Triols due to their nature do not possess photopolymerizable terminal groups and therefore cannot crosslink on their own which is why PCL derivatives have been developed [38
]. Like PCL Triol, poly(ethylene glycol) (PEG) itself cannot crosslink [41
] and has also been investigated for use as a plasticizer [42
] and in drug delivery applications [11
]. Martinez et al. [22
] conducted a study in which they added PEG 300 to the resin formulation and due to its inability to crosslink it would allow for a reduction in the intermolecular forces between the polymer chains, resulting in greater drug release.
Poly(ethylene glycol) (PEG) is a well-established synthetic polymeric material with PEG based hydrogels being well documented within the literature [29
] owing to the exceptional properties they afford, such as being nontoxic, biocompatible, and having notable tunability, in addition to being approved by the FDA for use in a variety clinical uses [51
]. PEGDA has been shown to be a beneficial polymer in drug delivery applications and was investigated by McAvoy et al. [53
] to evaluate its potential to fabricate implants via UV polymerization in the release of both small and large drug molecules with the former being released at a greater rate than that of the latter. This indicates that by the careful selection of drugs it may be feasible to manipulate and control the rate at which they are released from the fabricated dosage form.
The model drugs, paracetamol and aspirin, were found to easily dissolve in their respective photopolymer solutions with the colors of both solutions being clear and the 3DP dosage forms displaying the same color as in Figure 3
a,b. Although the concentrations of the model drugs were low the authors selected these concentrations based on previously published contributions with Goyanes et al. [20
] fabricating topical drug delivery systems with the incorporation of 2.00% w/w
salicylic acid the primary metabolite of aspirin. In addition to this, Wang et al. [41
] and Martinez et al. [6
] both fabricated dosage forms with the loading of paracetamol being 5.90% w/w
and 4.00% w/w
respectively. While the concentrations utilized do not have immediate clinical relevance, the formulations are intended to demonstrate the capability of the AM technique for bespoke batch production within a clinical setting. Despite this, the authors do believe that drug concentration could be increased, with Martinez et al. [54
] incorporating 10.00% w/w
of both model drugs used in this study, in addition to four other model drugs, to fabricate a multi-layered polypill via SLA. However, the effect on the increased drug concentration and thus decreased polymer concentration would need to be tested in order to identify how it would affect the printing process, curing process and drug release mechanism. Paracetamol and aspirin were identified as model drugs for this work, in part because of the difference in pKa values, 9.5 and 3.5 respectively [55
], of the two active ingredients. A rectangular shape was selected as the geometry to be printed, as shown in Figure 3
. However, 3DP has the capabilities of fabricating dosage forms of varied and complex shapes which has been demonstrated by Martinez et al. [6
] who evaluated the role that geometry plays in drug release from printed dosage forms.
As can be observed in Figure 3
a, it was possible to fabricate 28 tablets in one single print illustrating the potential the 3DP affords in the area of personalized medicine specifically if the patient requires one dosage per day meaning that it is possible to fabricate a month’s supply in one print. This finding has the possibility to revolutionize how pharmaceuticals are fabricated, and in particular when developing drug delivery methods for the treatment of rare diseases, as there is little to no waste from the printing process. Furthermore, through the manipulation of the dimensional design parameters of the dosage form, it may also be feasible to fabricate an amount greater than what was achieved in this study, 28 dosage forms, which would be advantageous if the patient was required to take more than one dose per day.
3.2. Spectroscopic Analysis of Photoinitiator
This study utilized Irgacure TPO which is a highly efficient photoinitiator that is used in the initiation of free radical polymerization [57
] with McDermott et al. [58
] reporting that it absorbs radiation in the region of 420–340 nm, making it ideal for use in the SLA process as the wavelength utilized by the SLA printer is 405 nm, as reported by Dizon et al. [59
]. The UV/Vis spectra of the photoinitiator Irgacure TPO was measured at three different concentrations, namely 0.10%, 0.50%, and 1.00%, and in two different solvents, 2-propanol and methanol, to establish if the photoinitiator has the capabilities of absorbing in the operating range of the SLA printer. As can be observed in the spectra, as shown in Figure 4
a,b, all three concentrations in both solvents absorb in the desired range, 405 nm. In both solvents, the TPO at a concentration of 0.10% showed a rather low absorption at 405 nm when compared to the other two concentrations, 0.50% and 1.00%. Furthermore, Arikawa et al. [60
] evaluated a number of photoinitiators and light curing units and concluded that both TPO and the violet-LED light was the most optimum in the polymerization of a mixture of bisphenol A-glycidyl methacrylate (Bis-GMA) and triethylene glycol dimethacrylate (TEGDMA).
Despite only a slight variance being apparent in the absorbance value between the concentrations, 0.50% and 1.00%, the latter concentration was selected for the study conducted. The rationale for selecting this concentration, 1.00%, is owing to the fact that if there is insufficient photoinitiator incorporated in the formulation this may result in surface cure problems. However, on the other hand, if too high a concentration is selected, high levels of surface crosslinking may occur which will prevent the UV light used in the printing process from effectively penetrating to the lower layers which may result in inadequate curing. Although the loading of 1.00% may seem high there are a number of publications in the literature utilizing higher concentrations. In a publication by Steyrer and researchers [61
], they illustrated the possibility of using a concentration loading of 1.18 wt% of Omnirad TPO-L (Ethyl (2,4,6-trimethylbenzoyl) phenylphosphinate) in the fabrication of 3D printed part on a 405 nm Digital Light Processing (DLP) printer, and although it may not be the same process used in this study it does, however, rely on a similar technology using UV light to cure the photopolymer. In addition to this, a concentration in excess of what is proposed in this study was utilized by Asikainen et al. [62
] in which they assessed lidocaine drug release from photo-crosslinkable PCL scaffolds fabricated by SLA technology.
However, the photoinitiator used in the studies by Steyrer and co-workers [61
] and Asikainen and researchers [62
] was an acylphosphine oxide derivative similar to that of Irgacure TPO used in this study. Furthermore, Bausch & Lomb Incorporated. filed a patent in 2010 utilizing a 1.50 wt.% of Lucirin TPO, now called Irgacure TPO, in the formulation of contact lenses [63
]. In addition to this, Martinez et al. [6
] also carried out a study, to explore the role that geometry plays in the release of drugs from printed tablets, using the same photoinitiator that is to be used within this study at a concentration of 2.00% w/w
. Additionally, both the photoinitiator, Irgacure TPO, and SLA printer, Formlabs Form 2, used in this study were also exploited in a study by Duan et al. [64
], albeit that the concentration of photoinitiator was higher in their study.
3.3. Fourier Transform Infrared Spectroscopy
The printed dosage forms were characterized in order to evaluate the efficiency of the SLA printing process as the IR spectral changes that occurred during the photopolymerization process could be evaluated. This evaluation is possible as the alterations, peak shifts, which occur in the spectra are indicative that bonding has occurred between the individual constituents present in the photopolymer solution, therefore suggesting that crosslinking has occurred within the samples [57
The double bond acrylate groups present at the end of the PEG chain in the PEGDA molecule is what gives rise for the polymer to undergo free radical polymerization in the presence of the TPO photoinitiator resulting in the fabrication of the dosage form. The UV light of the SLA printer activates the TPO, resulting in the TPO decomposing to generate two radicals, as previously shown in Figure 2
. The free radicals are then available to react with the PEGDA monomer in the photopolymer solution resulting in the opening of the C=C bond present in the PEGDA monomer thus allowing a polymer network to be formed [65
]. From conducting FTIR analysis of PEGDA (data not shown) the acrylate groups, C-H bending of the CH2
=CH–C=O and C–O–C stretching at 811 cm−1
and 1189 cm−1
respectively were identified with similar findings being reported elsewhere [66
a shows the results obtained from FTIR analysis of the paracetamol loaded dosage forms with the spectra for the aspirin samples being displayed in Figure 5
b. The FTIR spectra of all printed samples, paracetamol and aspirin, were similar to one another, however, due to the higher ratio of PCL Triol present within the composition, there was over-saturation of the notable peaks of PCL Triol (data not shown) found in printed samples.
As already stated, the higher ratio of PCL Triol within the photopolymer solution resulted in the spectra predominantly appearing as PCL Triol, with the broad peak ranging from 3394.19 to 3369.64 cm−1
in the samples being indicative of the free hydroxyl group (–OH) of the PCL Triol while the strong peak ranging from 1729.63 to 1726.05 cm−1
is characteristic of the stretching carbonyl (C=O) group of the molecule. It can also be observed that when both the 2.50% Para Resin and 2.50% Asp Resin is compared to their respective printed dosages there is the disappearance of peaks at 811 cm−1
, and 1189 cm−1
with similar findings being reported by Killion et al. [69
] and Burke et al. [66
] reporting that polymerization of a PEGDA derivative, PEGDMA, had occurred upon disappearance of the aforementioned peaks. This finding indicates that the SLA process undertaken within this study may be beneficial within the area of personalized medicine owing to its ability to cure the photopolymer solution and produce dosage forms of not only different active ingredients but also at different drug loading concentrations.
3.4. Thermal Characterization
The study of the thermal properties of the stereolithographic printed tablets and materials used in the fabrication of such tablets was evaluated by the utilization of DSC. Figure 6
shows the DSC thermograms of the materials used in the fabrication of the printed dosage forms. It can be observed that the photoinitiator, TPO, demonstrates a peak a melting peak around 92 °C, which corresponds to the melting point that is found in the materials technical data sheet [70
]. The DSC of PEGDA displays an endotherm which is clearly identifiable at 14 °C, which is similar to what Martinez and researchers [22
] found, establishing the melting point of PEGDA to be ca. 10 °C from the research that they conducted.
displays the DSC thermograms for paracetamol (first and second heat cycles) and the two printed dosage forms which contain paracetamol. The DSC for paracetamol, (a) first heat cycle, shows a sharp endothermic peak at 169 °C with this melting point being in agreement with that which is found in the literature [5
]. Furthermore, this melting point is indicative of paracetamol existing in its most stable form, that being the commercially marketed version, Form I (monoclinic) [72
]. This is further confirmed upon observing the second heat cycle (b) as the melting point has shifted to a lower temperature of 159 °C with Sacchetti [72
] noting similar results. Also, when the two heat cycles are compared, there is no cold crystallization peak noted in the first heat cycle, thus indicating that the paracetamol is present in its 100% crystalline form. However, upon observing the second heat cycle, it is apparent that the drug is 100% amorphous (Form II) with a glass transition temperature at 24 °C, with this finding been consistent with that of the literature [73
]. Furthermore, the broad exothermal peak between 75 and 100 °C, with peak maxima around 80 °C, is typical of cold crystallization occurring with Rengarajan and Beiner [73
] reporting results to support this. In addition to this, Rengarajan and Beiner [73
] also indicate that the Form II crystals melt at 157 °C with the results of this study finding the onset to be at 155 °C with a maximum melt peak at 159 °C, which according to the literature the melting temperature ranges from 154 to 160 °C which is characteristic of Form II, orthorhombic paracetamol [71
The overlaid DSC thermograms of aspirin and printed dosage forms containing aspirin can be observed in Figure 8
. The DSC for aspirin shows two melt peaks with the first endothermic peak being sharp and present at 144 °C (onset temperature 137 °C) with similar results being reported in the literature [75
]. However, the second endotherm was noted at 179 °C (onset temperature 177 °C) may be attributed to the degradation of aspirin. There are various publications within the literature, [78
], indicating that thermal degradation/decomposition of aspirin occurs in two processes with the first process ranging from 160 to 260 °C. Furthermore, thermogravimetric analysis (TGA) studies conducted by Al-Maydama et al. [81
] and de A. Silva et al. [82
] deduced that the temperature of maximum degradation (TDTG
) of aspirin occurred around 181 °C with further DSC studies conducted by de A. Silva et al. [82
] confirming this result.
As can be observed in both Figure 7
and Figure 8
, there was no endotherm observed for the dosages which contained either drug, paracetamol or aspirin, and either concentration, 2.50% or 5.00%, which is indicative of the active drug being entirely dissolved within the photopolymer solution. Similar results illustrating drug dissolution in the SLA photopolymer solution have been established with a number of publications [20
] citing comparable findings.
3.5. Tablet Characterization
In order to evaluate the effectiveness of SLA as a suitable means of fabricating tablets to be utilized in the area of personalized medicine, both the dimensional accuracy of the printing process and uniformity of tablet weight were evaluated. The tablets were fabricated with a high degree of repeatability with respect to their physical dimensions and this was apparent due to the paracetamol and aspirin printed dosage forms producing results close to target dimensions: Length 25.00 mm, width 10.00 mm, and height 5.00 mm, as shown in Figure 9
and Table 2
. However, some statistically significant differences were observed when the printed tablets were compared to that of their target dimensions. It was found that the 2.50% Para dosage forms displayed significant differences (p
< 0.05) in both their length and width when compared to that of their target dimensions. In addition to this, significant differences (p
< 0.0001) in both length and width were also observed for both the 2.50% Asp and 5.00% Asp dosage forms. Furthermore, there were no significant differences (p
> 0.05) noted for the 5.00% Para dosage forms in their length and width when compared to that of the target dimension as well as no significant differences (p
> 0.05) observed for any of the printed dosage forms irrespective of drug or drug loading in regard to the height of the fabricated dosage forms when compared to the expected dimension. The findings presented here are representative of what has previously been outlined in the literature. In two separate studies by Martinez et al. [22
], the authors assessed the dimensional accuracy of SLA printing as a means of fabricating dosage forms for drug delivery and concluded it was achievable to produce prints with good uniformity in physical dimensions.
Further to the dimensional analysis being conducted, weight uniformity among the printed tablets was also evaluated with the results presented in Figure 10
. Although the absolute weights of the two paracetamol formulations vary slightly from one another, all printed tablet weights ranged from 1324 to 1467 mg with no statistically significant difference (p
> 0.05) between the two formulations regarding tablet weight observed. However, it was found that the absolute weights of the two aspirin formulations did exhibit some variation, with the printed tablet weights ranging from 1566 to 1918 mg and the differences in tablet weight being significant (p
In addition to comparing the printed dosage forms containing the same drugs, the tablets containing the same drug loadings, 2.50% or 5.00%, were also evaluated with respect to uniformity of weight. The printed dosage forms which contained a drug loading of 2.50% showed a significant difference in weight (p < 0.0001). Similarly, when the drug loading was 5.00% in the printed tablets they also showed a significant difference (p < 0.0001).
Overall, the weight uniformity of all the printed tablets, paracetamol and aspirin, ranges from a lower range of 1324 mg (paracetamol) to an upper range of 1918 mg (aspirin) with significant differences between some tablets being shown as previously outlined with the exception of significant differences between 2.50% Para and 5.00% Para.
These findings reported here of dimensional accuracy and weight uniformity as well as the observations reported in the literature [22
] illustrate that AM and in particular the SLA process has the potential to be exploited in the fabrication of dosage forms for a variety of medical conditions. This is evident in the ability to print dosage forms comprised of various monomer(s)/photoinitiator(s)/drug concentrations and/or combinations of each which show excellent consistency in their dimensions with reference to the STL file that was imported into the software of the printer. Furthermore, the results obtained in this study indicate that it was feasible to fabricate a large batch of tablets in a single print, Figure 3
a, with 28 dosage forms being produced. In addition to this finding, it should also be noted that a greater number of tablets could be attained through manipulation of the software which would be advantageous when treatment is required more than once per day.
3.6. Dissolution Analysis
Drug release studies were conducted for all of the printed dosage forms with the results shown in Figure 11
. As can be observed both the paracetamol and aspirin loaded dosage forms exhibited sustained drug release which was maintained for the duration of the study, 24 h.
Looking at Figure 11
, it can be observed that the printed paracetamol dosage forms displayed sustained release of the active drug over the 24 h testing period. It was found that the 2.50% Para exhibited a maximum release of 84.11% with the 5.00% loading achieving a release of 88.16% for the test duration. Over the test period it was found that upon comparison of the two different drug loadings there was statistical significance (p
< 0.0001) up to the two hour time point after which no statistical significance (p
> 0.05) was noted until the last and final time point, 24 h, (p
As with the paracetamol dosage forms, the aspirin dosage forms also exhibited a sustained rate of release from the 3DP matrices. However, there was a difference in the maximum release after 24 h with the 2.50% Asp having a maximum release rate of 95.85% after the 24 h period with the higher loading, 5.00% Asp, achieving a higher maximum release rate of 113.30% after the same time. However, although the 5.00% Asp dosage form released a greater amount of drug after 24 h, this is believed to have been due to the higher loading of the drug, and therefore less polymer within the polymer matrix. Furthermore, the rate at which aspirin was released from the printed dosage forms was not statistically significant (p > 0.05) for time points up to 4 h.
The results presented in this study differ from that reported by Wang and researchers [41
] who describe that by altering the ratio of PEGDA to PEG it was possible to modulate drug release, of paracetamol and 4-amino salicylic acid, from the printed dosages with a low percentage of PEGDA (35%) showing faster release rates from that of their higher counterparts. The percentage of PEGDA which was utilized in this study was lower than what was used in the study by Wang et al. [41
] with slower release rates been observed with a lower PEGDA content. It was found in the study by Wang and researchers [41
] that at the 8 h (480 min) time point that close to 95% of paracetamol was released while at the same point in this study there was only 53% and 56% released from the 2.50% and 5.00% respective paracetamol loaded dosage forms. However, one explanation for this is the incorporation of PEG in the matrix of the dosage form as due to its hydrophilic nature and its inability to crosslink would allow for a reduction in the intermolecular forces between the polymer chains resulting in greater drug release [22
]. Additionally, the results also conflict with what was presented by Vehse et al. [87
] who incorporated various concentrations of aspirin into PEGDA matrices via a similar technology albeit a different wavelength, and found that more than 95% of aspirin was released within the first 3 h regardless of the drug loading concentration.
The lower release rates established in this study may be attributed to the higher PCL Triol content of the resin composition used in the formulation, with the PCL Triol influencing drug release due to the hydrophobic nature it affords. Asikainen et al. [62
] noted that an increase in hydrophobic networks will result in slower drug release. Furthermore, the addition of low molecular weight molecules added to polymer matrices have been reported to have the ability to modulate drug release due to a reduction in the polymer-polymer interactions as a result of the plasticization effect [32
]. Kanis et al. [34
] conducted a study evaluating the effect of PCL Triol content imparted on drug release from a poly(ethylene-co-methyl acrylate) matrix and found that by increasing the amount of PCL Triol resulted in a lower amount of drug being released from the matrix. Therefore, it could be possible to control and manipulate the rate at which drug is released from the dosage form by adjusting the ratio of PCL Triol to PEGDA in the formulation, which would be advantageous if immediate release was a requirement of the dosage form.
3.7. Scanning Electron Microscopy
Scanning electron microscopy (SEM) images of selected paracetamol and aspirin printed dosage forms are presented in Figure 12
From observing the images on the surface of the paracetamol loaded dosage forms, Figure 12
a, it can be seen that there are no visible voids present indicative of the photopolymer resin curing to a high degree which was further confirmed via FTIR analysis, with the disappearance of the peaks at 811 cm−1
and 1189 cm−1
associated with polymerization occurring [67
]. However, in the images displaying the cross-section of the 2.50% Para dosage form, Figure 12
b, it appears that there are notable crystals present which may be TPO and/or paracetamol, suggesting that some of the crystals may not have dissolved in their entirety in the photopolymer solution.
The images obtained for the aspirin dosage forms however, Figure 12
c,d, appear different to those of the paracetamol dosage forms. The image showing the surface of the 2.50% Asp dosage form, Figure 12
c, the surface appears coarse in comparison to that of the paracetamol dosage form which may be as a result of undissolved TPO and/or aspirin settling on the surface of the print with further analysis that is beyond the scope of this work being required to ascertain this finding. In the cross-sectional images of the 2.50% Asp dosage forms, Figure 12
d, the layering of the printing process/laser pass can be observed which demonstrates that polymerization of resin occurred with the binding of the resin in a layer-by-layer fashion as would be expected in the SLA process. However, in addition to this, in the 2.50% Asp dosage form there is evidence of voids present, indicating that incomplete curing may have occurred in these areas. Furthermore, there was also particles observed in the 2.50% Asp dosage form which may be present as a result of the TPO and/or aspirin not fully dissolving into the photopolymer solution.
The results found from conducting SEM demonstrate that SLA has the potential to fabricate dosage forms from photocurable resins, with little to no voids being observed on the surface of the dosage forms being indicative of a high level of curing. However, upon closer inspection and viewing the cross-sectional images of the dosage forms, it was found that there may have been incomplete dissolution or agglomeration occurring due to the inhomogeneous distribution of either drug and/or photoinitiator within the resin as crystals were noted. This may be a potential drawback to SLA as an AM technique for drug delivery applications with these findings requiring further investigation in order to elucidate if the observed particles are that of drug and/or photoinitiator.