International Journal of Pharmaceutics 564 (2019) 98–105

Contents lists available at ScienceDirect

International Journal of Pharmaceutics

journal homepage: www.elsevier.com/locate/ijpharm

3D printed oral theophylline doses with innovative ‘radiator-like’ design: T

Impact of polyethylene oxide (PEO) molecular weight
Abdullah Isreba, Krzysztof Bajb, Magdalena Wojszc, Mohammad Isrebd, Matthew Peake,
⁎
Mohamed A Alhnanf,
a
School of Pharmacy and Biomedical Sciences, University of Central Lancashire, Preston, Lancashire, UK
b
Faculty of Pharmacy, Medical University of Lodz, Lodz, Poland
c
Faculty of Pharmacy with the Laboratory Medicine Division, Medical University of Warsaw, Warsaw, Poland
d
School of Pharmacy, University of Bradford, Richmond Road, Bradford, UK
e
Paediatric Medicines Research Unit, Alder Hey Children’s NHS Foundation Trust, Liverpool, UK
f
Institute of Pharmaceutical Science, King's College London, London, UK

A R T I C LE I N FO A B S T R A C T

Keywords: Despite the abundant use of polyethylene oxides (PEOs) and their integration as an excipient in numerous
Personalised medicine pharmaceutical products, there have been no previous reports of applying this important thermoplastic polymer
Additive manufacturing species alone to fused deposition modelling (FDM) 3D printing. In this work, we have investigated the manu-
Complex structures facture of oral doses via FDM 3D printing by employing PEOs as a backbone polymer in combination with
Tablets
polyethylene glycol (PEG). Blends of PEO (molecular weight 100 K, 200 K, 300 K, 600 K or 900 K) with PEG 6 K
Patient-specific
Structural design
(plasticiser) and a model drug (theophylline) were hot-melt extruded. The resultant filaments were used as a feed
for FDM 3D printer to fabricate oral dosage forms (ODFs) with innovative designs. ODFs were designed in a
radiator-like geometry with connected paralleled plates and inter-plate spacing of either 0.5, 1, 1.5 or 2 mm. X-
ray diffraction patterns of the filaments revealed the presence of two distinctive peaks at 2θ = 7° and 12°, which
can be correlated to the diffraction pattern of theophylline crystals. Blends of PEO and PEG yielded filaments of
variable mechanically resistance (maximum load at break of 357, 608, 649, 882, 781 N for filament produced
with PEO 100 K, 200 K, 300 K, 600 K or 900 K, respectively). Filaments of PEO at a molecular weight of
200–600 K were compatible with FDM 3D printing process. Further increase in PEO molecular weight resulted in
elevated shear viscosity (> 104 Pa.S) at the printing temperature and hindered material flow during FDM 3D
printing process. A minimal spacing (1 mm) between parallel plates of the radiator-like design deemed essential
to boost drug release from the structure. This is the first report of utilising this widely used biodegradable
polymer species (PEOs and PEG) in FDM 3D printing.

1. Introduction Among other commercially available technologies, fused deposition

modelling (FDM) 3D printing offers major advantages, including the
Through recent advances in pharmacogenetics the relationship be- low cost of the printer, the absence of finishing steps and the lack need
tween an individual’s genome, their genetic predisposition to disease for powder facilities. These properties position FDM 3D printing as a
and their response to specific medications is increasingly understood very attractive platform for small-scale individualising for solid dosage
(Hamburg and Collins, 2010). With an increased focus on patient- forms. Recently, several examples of the use of FDM 3D printing for
centred and stratified treatment, there is a growing need for a techno- production of immediate, delayed and extended drug release have been
logical solution to provide individual patients with reliable and safe reported (Palo et al., 2017; Goyanes et al., 2015; Tagami et al., 2017;
personalised dosage forms. In the last few years, additive manu- Pietrzak et al., 2015; Sadia et al., 2016; Skowyra et al., 2015). The
facturing has been proposed as an alternative platform for on-demand technology proved effective at accurately titrating coumarin doses in
production of personalised dosage forms with significant ability to animals (Arafat et al., 2018) and extended drug release in gastro-re-
tailor the size, shape, dose as well as drug release pattern (Alhnan et al., tentive systems (Li et al., 2018).
2016; Palo et al., 2017; Prasad and Smyth, 2016). For the pharmaceutical industry to make a full use of 3D printing, it

⁎
Corresponding author at: Institute of Pharmaceutical Sciences, King's College London,150 Stamford Street, London SE1 9NH. Tel.: +44 (0)20 7848 7265.
E-mail address: [email protected] (M.A. Alhnan).

https://doi.org/10.1016/j.ijpharm.2019.04.017
Received 13 November 2018; Received in revised form 3 April 2019; Accepted 6 April 2019
Available online 08 April 2019
0378-5173/ © 2019 Elsevier B.V. All rights reserved.
A. Isreb, et al. International Journal of Pharmaceutics 564 (2019) 98–105

is essential to adapt pharmaceutical grade polymers for FDM 3D

Compatibility with the gears of FDM 3D printer head

printing. Previous studies have used cellulose, methacrylate, acrylic
acid or PVP derivatives to produce solid dosage forms (Awad et al.,
2018). PEO is one of the most commonly used polymers in pharma-
ceutical industry. PEO is commercially available between 100 K and
10,000 K g/mole and has been extensively used for oral and parental
formulations (Gullapalli and Mazzitelli, 2015). PEOs have been com-
monly used to produce extended release tablets in powder compression
Moroni and Ghebresellassie, 1995), hot melt extrusion (Zhang and
McGinity, 1999) and in buccal tablets (Apicella et al., 1993). However,
limited reports are available applying this extensively used polymer
species to FDM 3D printing. In rare examples, PEO was used for for-

Compatible
Compatible
Compatible
Too fragile
mation of thin oral film in combination with other additives (Ehtezazi

Fragile
et al., 2018), or as an additive to methacrylate polymer for 3D printing
of tablets (Alhijjaj et al., 2016).
In order for a filament to be compatible with the FDM 3D printing
process, it requires critical mechanical and rheological criteria
(Nasereddin et al., 2018). Previous studies have linked a filament’s 3D
printing compatibility with the rheological properties of the backbone

FDM 3D Printing Temp. (°C)

polymers: poly methacrylate (Sadia et al., 2016), PVA (Boetker et al.,
2016; Pereira et al., 2019) and PVP-VA (Boetker et al., 2016;

Composition, processing temperatures and FDM 3D printing compatibility of theophylline filament including PEO of different molecular weights.
Fuenmayor et al., 2018). The availability of PEOs at different molecular
weight grades provides the opportunity to test the impact of polymeric
molecular weight and rheological flow properties of a single polymer.
In this work, we have investigated the fabrication of oral doses via

NA**
NA*
105
110
145
FDM 3D printing by employing PEOs as a backbone polymer in com-
bination with PEG as a plasticiser. We assessed the impact of polymer
molecular weight on the mechanical properties of the resultant fila-
ments and their rheological properties. We have also tested the effect of
an innovative radiator-like design of the solid dosage form on the ac-
ExtrusionTemp. (°C)

celeration of drug release patterns.

2. Materials and methods

2.1. Materials
60
65
70
80
80

Theophylline was supplied by Acros Organics (UK). Polyethylene

glycol (PEG 6000) and all grades of polyethylene oxide (PEO) were
** Filaments were incompatible with FDM 3D printer due to immediate nozzle blockage.
* Filaments were incompatible with FDM 3D printer due to frequent filament breakage.

purchased from Sigma-Aldrich (Dorset, UK).

ProcessingTemp. (°C)

2.2. Preparation of filaments using hot melt extrusion (HME)

Filaments were prepared by mixing polyethylene oxide (PEO mo-

HME

lecular weight of 100 K, 200 K, 300 K, 600 K, or 900 K), Polyethylene

60
65
70
80
80

Glycol (PEG 6 K) and theophylline (Table 1). The mixtures were ex-
truded using a Thermo Scientific HAAKE MiniCTW hot melt extruder
(Karlsruhe, Germany) after mixing inside the extruder for 5 min at a
temperature range of 60–80 °C (Table1) at 35 rpm using 1.5 mm nozzle.
30:35:35
30:35:35
30:35:35
30:35:35
30:35:35

2.3. Oral dosage forms (ODFs) design and printing

100 K
200 K
300 K
600 K
900 K

ODFs were designed using Autodesk® 3ds Max Design 2016 software
PEO
PEO
PEO
PEO
PEO

version 18.0 (Autodesk, Inc., USA). In the CAD design, the radiator-like
desgins were structured with increasing inter-plate spacing of 5, 10, 15
6 K:
6 K:
6 K:
6 K:
6 K:

or 20 mm whilst the overall dimensions of the design were maintained

Theophylline:PEG
Theophylline:PEG
Theophylline:PEG
Theophylline:PEG
Theophylline:PEG

within the volume of 20 × 10 × 6 mm. The templates were then im-

Composition

ported into the 3D printer software in a stereolithography (.stl) file

format. The previously extruded filaments were fed into the FDM 3D
printer equipped with 0.4 mm nozzle size and MakerWare Version
2.4.0.17 (Makerbot Industries, LLC, USA). ODFs were printed using
modified settings of the software as described earlier in our previous
work (Skowyra et al., 2015):Replicator 2X; type of filament: PLA; re-
solution: standard; temperature of building plate: 40 °C; speed of ex-
100 K
200 K
300 K
600 K
900 K
Filament

truder 50 mm/s while extruding and 150 mm/s while traveling; infill:
Table 1

100%; height of the layer: 200 µm. The temperature of the nozzle for
Fil
Fil
Fil
Fil
Fil

each filament is specified in Table 1.

99
A. Isreb, et al. International Journal of Pharmaceutics 564 (2019) 98–105

2.4. Thermal analysis Vernier micro-caliper for various sections and the average (c.a. 1.8 mm)
was programmed into the software. The deformation rate (extension)
Thermal decomposition profiles for PEOs as both received and ex- was set to 20 mm/min and the data were collected every 50 ms. A sand
truded filaments were measured using a TA Q500 Thermogravimetric paper was used to prevent the slipping of the sample from the clamp.
Analyzer TGA (TA Instruments, Elstree, Hertfordshire, UK). Samples Samples that showed signs of slipping from the clamp were rejected and
with an average weight of 10 mg were measured from 25 °C to 500 °C all samples were measured in triplicate. A stress strain graph was
with a heating rate of 10 °C/min and a nitrogen gas purge of 40/60 mL/ plotted for each sample and the breaking stress was measured.
min for sample/furnace respectively. The thermal behaviour of these
samples was measured using a TA Q2000 Differential Scanning 2.10. Drug contents and in vitro drug release studies
Calorimeter (DSC) (TA Instruments, Elstree, Hertfordshire, UK).
Samples (5 mg) were prepared in aluminium standard pans (40 µL) and For assessment of theophylline contents, oral doses were dissolved
sealed with pin-holed lid. Each sample was heated from −10 to 255 °C in 500 mL of deionised water and were stirred continiuously for one
at 10 °C/min under a nitrogen purge of 50 mL/min. Data from TGA and hour at 40 °C until complete dissolution. Samples were filtered through
DSC were analysed using a TA 2000 analysis software (TA Instruments, a 0.22 μm Millex-GP syringe filter (Merck Millipore, USA) and the
Elstree, Hertfordshire, UK). All measurements were carried out in tri- concentration of the drug was determined using a Jenway
plicate. Spectrophotometer (Bibby Scientific Ltd, UK) at λ max of 272 nm
(n = 3).
2.5. X-ray diffractometry (XRD) To study in vitro theophylline release for 3D printed ODFs, an AT 7
Smart USP II dissolution test apparatus (Sotax, Switzerland) was used. A
An X-ray powder diffractometer, D2 Phaser with Lynxeye (Bruker, dissolution medium of 900 mL 0.1 M HCl (pH 1.2) at 37 ± 0.5 °C with
Germany) was used to assess the physical form of theophylline, PEO, a paddle speed of 50 rpm was used for 2 h. Each experiment was carried
PEG and drug loaded filaments. Samples were scanned from (2θ) = 5° out in triplicate. Samples were collected at 5 min intervals and drug
to 50° using 0.01° step width and a 1 s time count. The X-ray wave- concentration was determined using UV/VIS spectrophotometer (PG
length of 0.154 nm was used using a Cu source and a voltage of 30 kV. Instruments Limited, UK) at the wavelength of 272 nm and path length
The divergence slit was 1 mm and the scatter slit 0.6 mm. Filament of 10 mm and outcome data were analysed using IDISis software 2012
emission was 10 mA using a scan type coupled with a two theta/theta (Automated Lab, UK).
scintillation counter over 60 min.
2.11. Statistical analysis
2.6. Hansen solubility parameter
The data were analysed by one-way ANOVA using SPSS Software
Hansen solubility parameters for the polymers and the drugs were (22.0.0.2). The level of attributed significance for comparisons were as
calculated using HSPiP software (version 5.0.08). follows: p > 0.05 not significant; p ≤ 0.05 significant.

2.7. Scanning electron microscopy (SEM) 3. Results and discussion

The topography of the drug-loaded filaments and the 3D printed Polyethylene glycol (PEG) and polyethylene oxide (PEO) are two of
ODFs were examined using Quanta-200 SEM microscope at 20 kV. the most widely used excipients in pharmaceutical products. Both
Samples were coated under vacuum with a gold coater JFC-1200 Fine products are also used in other healthcare applications. Both polymers
Coater (Jeol, Tokyo, Japan). In addition, photographs of ODFs were are also biodegradable and suitable to be used as a polymeric bioma-
collected a Canon EOS-1D Mark IV (Canon Ltd, Japan). terial in tissue scaffolding (Bliley and Marra, 2015). PEGs are con-
sidered a safe choice to prepare hydrogel sealant for patients under-
2.8. Rheology studies going surgery (Cosgrove et al., 2007) and are also used in the
manufacturing of 3D porous scaffolds (Husár et al., 2014). Optimisation
A shear Physica MCR 501 rheometer (Anton Paar, Germany) was of pharmaceutical solid dosage forms produced by FDM 3D printing
used in oscillation mode with a parallel plate configuration (plate requires a suitable and compatible polymer backbone for the feed fi-
diameter = 25 mm). The gap between the plate and the base was set at lament. Initially, PEGs were first assessed producing feed filaments for
0.5 mm. Amplitude sweep test was performed to determine the linear FDM 3D printing (as a backbone polymer). However, the hot melt ex-
viscoelastic region (LVR). Afterwards, frequency sweep tests were trusion process only yielded easily breakable PEG-based filaments
performed at a strain amplitude of 1% (Well within the LVR region) and which lacked the required rheological and mechanical properties to
an angular frequency range from 100 to 0.1 rad/s. Each sample was enable for the use of PEGs in FDM 3D printing of solid dosage forms
tested at three temperatures; 100, 110 and 140 °C. The readings (n = 6) (data not shown). Therefore, a higher molecular weight thermoplastic
were recorded for each frequency decade (18 points in total). The test polymer (PEO), was used for its mechanical and rheological properties
was only carried out after the normal force recorded by the device while PEG was added as a plasticiser to facilitate the material flow and
dropped below 1 N, which indicates that the polymer is in a relaxed pore former to accelerate drug release from the dosage form produced
state. Power law fit was used in the linear shear thinning area of the by FDM 3D printing.
obtained rheological data to measure the shear-thinning index (n ). The thermal properties of PEOs of different molecular weights
Elastic (G′) and viscous (G″) moduli as well as complex viscosity data (100 K-900 K) were shown to be stable at < 150 °C (Fig. 1A). In addi-
were recorded and plotted against the angular frequency at each tem- tion, PEO revealed a minimum moisture content with a weight loss
perature. of < 2% at 120 °C. The polymer showed no significant change in
thermal degradation following the compounding into a filament with
2.9. Tensile strength studies the addition of PEG and theophylline via HME extrusion (Fig. 1B). Pure
polymers melting points were observed above 66–69 °C (data not
A tensile strength testing system 5568 (Instron, Buckinghamshire, shown) (Beech and Booth, 1970). However, the compounded filament
UK) was used to measure the breaking stress for filaments with irregular produced in this study showed slightly lower melting points (in the
geometry with an average diameter of approximately 1.8 and 10 mm range of 62–65.9 °C), which could be attributed to the addition of a
gauge length. The diameter of the samples was measured using a lower melting point additive (PEG) (Fig. 1C). Thermal profiles also

100
A. Isreb, et al. International Journal of Pharmaceutics 564 (2019) 98–105

Fig. 1. TGA thermal decomposition profile of: A) raw PEO powder with molecular weights (100 K, 200 K, 300 K, 600 K, and 900 K), B) hot melt extruded filaments
containing 30% theophylline, 35% PEG 6 K, and 35% PEO (100 K, 200 K, 300 K, 600 K, and 900 K), and C) DSC thermographs of corresponding filaments.

illustrated that theophylline was crystalline within the polymer matrix

with the appearance of theophylline melting endotherm known to be at
∼240 °C (Hock et al., 2008).
XRD patterns confirmed the crystallinity status of PEO 200 K and
PEG 6 K with the presence of intensity peaks at 2θ = 19.1° and 23.2° the
appearance of these peaks in the pattern of HME compounded filament
suggests that polymers remained crystalline. The diffraction patterns of
extruded filaments also revealed diffraction peaks at 2θ = 7° and 12.9°
(Fig. 2). The later peaks are characteristic peaks in the diffraction
pattern of theophylline (Pietrzak et al., 2015). This confirmed the
crystalline structure of theophylline within the polymeric matrix. The
diffraction patterns of filaments produced with other molecular weight
PEOs (100 K, 300 K, 600 K and 900 K), also revealed the presence of
crystalline theophylline (Supplementary data, Figs. S1–4).
The Hansen solubility parameter data of the PEO and PEG blend and
the drug are shown in Table 2. The difference in solubility parameter
between the PEO and PEG blend and the drug (Δδ = 7 MPa1/2), in-
dicated a minimal miscibility between these molecules and predicted
the presence of theophylline as a solid suspension within PEG/PEO
polymeric matrix.
The impact of molecular weight on mechanical properties of HME
compounded filaments was assessed using the tensile strength test
(Fig. 3A). HME compounded filaments including PEO of 100 K mole-
cular weight showed the least maximum load before break (357 N)
(p < 0.05) and were deemed too fragile. As the filament breaks in- Fig. 2. Representative XRD diffraction patterns of raw theophylline, PEG 6 K,
stantly upon the application of gear pressure in the FDM 3D printer’s raw PEO 200 K, and hot melt extruded filament containing 35% PEO200K, 35%
PEG 6 K, and 30% theophylline (for other grades, see Fig. S1 in Supplementary
head. HME compounded filaments including PEO of 200 K molecular
Data).
weight were able to be loaded through the gears of the FDM 3D printer
head. However, frequent breakage of the filament due to the pressure of
the gears interrupted the printing process and resulted in printing increased with longer polymer chains (Husken and Gaymans, 2009). On
failure. When HME compounded filaments containing PEO of higher the other hand, Young modulus of PEO 100 K based filament reveal
molecular weight (300 K, 600 K and 900 K), the filaments were able to more brittle behaviour in comparison to filaments produced with
withstand higher tension (Fig. 3A). The maximum load at break steadily higher molecular weight PEO (Fig. 3B). The increased plasticity of HME

101
A. Isreb, et al. International Journal of Pharmaceutics 564 (2019) 98–105

Table 2 because increasing temperature above 150 °C is likely to accelerate PEO

Hansen solubility parameter in MPa1/2and components due to theophylline and degradation (Crowley et al., 2002). It can be deduced that a complex
PEO/PEG. viscosity of approximately < 8000 Pa.S is necessary to achieve suffi-
δD δP δH HSP cient material flow from the FDM 3D printer hot nozzle and successful
completion of FDM 3D printing.
Theophylline 19.7 15.4 10.5 27.1 The viscoelastic properties of the filaments produced with PEOs of
PEO, PEG 17 10 5 20.3
different MW were characterised through the measurement of the sto-
Where δD, δP and δH are the dispersion, polar, and hydrogen components of rage G‘ and loss modulus G″ (Fig. 6). In general, increasing the tem-
solubility parameter and (HSP) Hansen solubility parameter. perature led to a decrease in both storage modulus G‘ and loss modulus
G″ across different molecular weights. Filaments containing PEO 100 K
compounded filaments containing higher molecular weight PEO also were noticed to be in a terminal flow zone as G″ > G′. A higher PEO
allows the filament to withstand more pressure from the gears of the molecular weight in the filament resulted in less liquid-like flow and a
head of the FDM 3D printer and mitigates the risk of filament breakage. more elastic behaviour as the polymer was approaching crossover
This increase in the strength of the filament can be related to previous point. Following extrusion from the nozzle of the 3D printer, the fila-
observations of the reduced mobility due to the entanglement of the ment loses its microstructure and conforms to the architecture dictated
amorphous parts of the polymeric chains associated with an increase in by the CAD design and slicing engine. This behaviour can be advanta-
the chain length (Kennedy et al., 1994). geous in FDM 3D printing as it provides a wide variety of molecular
During the FDM 3D printing process, the filament passed through weights to select from while maintaining the same release profile. This
lead to a hot channel that terminates in a nozzle and while the path is observation needs to be repeated and validated with other drugs that
narrowed from 1.75 to 0.4 mm (nozzle diameter), the filament experi- may interact with PEO.
ences an increase from room temperature to the printing temperature Unlike regular caplet design, where filaments are not only fused
(110–145 °C). Therefore, it is essential to study the rheological beha- with lower and upper layers, but also with side printed layers, the ra-
viour of the filament compositions at the temperature of the printing diator-like design only allows fusion with upper and lower layers,
nozzle. Hence, complex viscosity under various angular frequency at leading to potentially different mechanical behaviour to solid caplet
two representative printing temperatures (110 and 145 °C) were per- design. However, it was not possible to measure the tensile strength of
formed (Fig. 4). Complex viscosity of a polymer is a temperature-de- oral doses due to their thin structure, where weak clamping point of the
pendent material property (Wissbrun, 1980). Despite the similarity of structures deemed it unsuitable for the test.
the melting points across all PEO grades, the printability of each fila- When theophylline release from capsule-shaped tablets with PEO
ment using FDM 3D printing was dependent on the temperature of the 600 K produced by FDM 3D printing was assessed (Supplementary data,
3D printer temperature (Table 1). The lower complex viscosities of PEO Fig. S2), a slow release profile was observed. It is likely that the drug is
100 K (539.8 Pa.S) and 200 K (1385.31 Pa.S) based filaments suggest released though erosion of the polymeric matrix and diffusion me-
possible flow from the hot nozzle of the 3D printer (Fig. 5). However, it chanisms (Shojaee et al., 2015). The polymer-rich structure of the ca-
was not possible to physically test 3D printing using these filaments due plet hindered drug release. The fast hydration of PEO/PEG based tablets
to their incompatibility with the gears of the 3D printer’s head (see produced by powder compression was reported to yield a gel-layer
above). However, HME compounded filaments including PEO of higher upon introduction to dissolution medium that significantly prolongs
molecular weights allowed consistent flow from the hot nozzle at a drug release (Maggi et al., 2000). In fact, PEO/PEG blends have been
printing temperature of 110 and 145 °C for PEO 300 K and 600 K re- devised to produce tablet with extended release over 12–24 h (Kojima
spectively (Fig. 5). The complex viscosity of these filaments was in the et al., 2008). In such matrix systems, drug release is dependent on the
range of 9000 and 10,000 at the corresponding temperature at 1% rate of polymer dissolution (Apicella et al., 1993), which regulates the
angular viscosity. Filament containing higher molecular weight PEO pattern of drug release and often yields a zero order pattern (Kim,
(900 K) was observed to have a high complex viscosity (> 22610 Pa.S) 1998).
and was associated with restricted materials flow in the nozzle of the 3D In order to accelerate drug release from PEO matrix, an alternative
printer and obstructed the printing of this particular filament. Further novel design approach of a radiator-like architecture was evaluated
increase in nozzle temperatures (up to 220 °C), did not improve mate- (Fig. 7). The proposed geometry allows 7–8-fold increase in surface-to-
rial flow of this specific HME compounded filament. This may be mass ratio of the structure (Table 3). Moreover, the design facilitates
water penetration and drug permeation from the PEO matrix by

Fig. 3. Tensile strength data: A) maximum load at break and B) Young Modulus for hot melt extruded filaments containing 30% theophylline, 35% PEG 6 K, and 35%
PEO (100 K, 200 K, 300 K, 600 K, and 900 K).

102
A. Isreb, et al. International Journal of Pharmaceutics 564 (2019) 98–105

Fig. 4. Shear index for filaments containing 30% theophylline, 35% PEG 6 K, and 35% PEO (100 K, 200 K, 300 K, 600 K, and 900 K) at 110 and 145 °C.

Fig. 5. Shear rheometer data of complex viscosity for filaments containing 30% theophylline, 35% PEG 6 K, and 35% PEO (100 K, 200 K, 300 K, 600 K, and 900 K) at
A) 110 and B) 145 °C.

Fig. 6. Shear rheometer data of storage modulus and loss modulus for filaments containing 30% theophylline, 35% PEG 6 K, and 35% PEO (100 K, 200 K, 300 K,
600 K, and 900 K) at A) 110 and B) 145 °C.

minimizing the thickness of gel-layer with the use of low-thickness 20 × 10 × 6 mm, the number of plates has decreased and resulted in
plates. Four designs with identical overall dimensions but with in- lower printed mass and dose (Fig. 7, Table 3). A minimum spacing of
creasing spaces (0.5, 1.0, 1.5 and 2.0 mm) between the design plates 1 mm was deemed essential to accelerate drug release from the struc-
were tested. With increasing inter-plate spacing within the dimensions ture and meet USP criteria for immediate release products (Fig. 8A).

103
A. Isreb, et al. International Journal of Pharmaceutics 564 (2019) 98–105

Fig. 7. (A1) Rendered image and (A2) photograph of radiator-like design. (B1) Top view, (B 2) side view and (B3) photograph of radiator-like doses based on
theophylline:PEG 6 K:PEO 600 K 30:35:35.

Table 3
Dimensions, volume of CAD designs, mass and surface/mass ratios of caplet and radiator-like structures generated by FDM 3D printing.
Design CAD design dimensions (mm) CAD design volume CAD design surface Mass (mg) Surface /mass ratio Drug contents (mg)
(mm3) (mm2) (mm2/mg)
X Y Z

Caplet 12 4.75 4.36 188.33 184.07 172.78 ± 4.8 1.07 51.7 ± 2.2
Radiator-2 mm spaced 20 10 5.99 141.47 1464.39 187.7 ± 3.9 7.80 52.4 ± 1.7
Radiator-1.5 mm spaced 19.7 10 5.99 170.46 1705.64 196.2 ± 5.9 8.69 53 ± 0.7
Radiator-1 mm spaced 19.9 10 6.03 231.35 2211.34 247.86 ± 3.5 8.92 74.5 ± 0.6
Radiator-0.5 mm spaced 19.4 10. 5.99 262.16 2557.66 290 ± 7.9 8.82 80.1 ± 0.7

Fig. 8. In vitro release pattern of: A) 0.6, 1.0, 1.5, and 2.0 mm spaced radiator-like 3D printed ODFs containing 30% theophylline, 35% PEG 6 K, and 35% PEO 600 K,
and B) ODFs prepared using filaments composed of 30% theophylline, 35% PEG 6 K, and 35% PEO 600 K.

Similar drug release was obtained within the FDM 3D-printable range designs. It is possible that such swelling in the 0.5 mm-spaced design
of PEOs (200 K-600 K) (Fig. 8B). Following introduction to the dis- resulted in plate adhesion, leading to reduction of contact surface area
solution apparatus, the PEO matrix hydrates and swells leading to with the dissolution medium and hence slowing drug release. The paper
significant growth in the thickness of the radiator plate. Despite similar provides a unique example of how 3D printing and novel design ap-
surface-to-mass ratio of these oral dose designs (Table 3), the 0.5 mm proach can significantly alter the release profile of the same formula-
spaced design appeared to be slower in comparison with the rest of tion. The use of radiator-like design maximised interaction with

104
A. Isreb, et al. International Journal of Pharmaceutics 564 (2019) 98–105

dissolution medium and prevented the formation of thick permission Boetker, J., et al., 2016. Modifying release characteristics from 3D printed drug-eluting
gel layer, which will slow down drug release. In the future, such design products. Eur. J. Pharm. Sci. 90, 47–52.
Cosgrove, G.R., et al., 2007. Safety and efficacy of a novel polyethylene glycol hydrogel
approach will help to personalise the release profile without the need to sealant for watertight dural repair. J. Neurosurg. 106 (1), 52–58.
modify the formulation. Crowley, M.M., et al., 2002. Stability of polyethylene oxide in matrix tablets prepared by
hot-melt extrusion. Biomaterials 23 (21), 4241–4248.
Ehtezazi, T., et al., 2018. The application of 3D printing in the formulation of multi-
4. Conclusion layered fast dissolving oral films. J. Pharm. Sci. 107 (4), 1076–1085.
Fuenmayor, E., et al., 2018. Material considerations for fused-filament fabrication of solid
This work demonstrates the effect of PEO molecular weight on the dosage forms. Pharmaceutics 10(2).
Goyanes, A., et al., 2015. 3D printing of modified-release aminosalicylate (4-ASA and 5-
compatibility of HME compounded filaments for FDM 3D printing. A ASA) tablets. Eur. J. Pharm. Biopharm. 89, 157–162.
molecular weight of PEO between 300 K and 600 K was shown to have Gullapalli, R.P., Mazzitelli, C.L., 2015. Polyethylene glycols in oral and parenteral
optimal mechanical and rheological properties for the FDM 3D printing formulations–A critical review. Int. J. Pharm. 496, 219–239.
Hamburg, M.A., Collins, F.S., 2010. The path to personalized medicine. N. Engl. J. Med.
process. A lower molecular weight of PEO (100 K–200 K) yielded me-
363 (4), 301–304.
chanically incompatible HME compounded filaments and a larger mo- Hock, C., et al., 2008. Melting-point depression by insoluble impurities: a finite size effect.
lecular weight of PEO (900 K) contributed to significantly high complex Phys. Rev. Lett. 101 (2), 023401.
viscosity and inhibited material flow. The use of a relatively low Husár, B., et al., 2014. Photopolymerization-based additive manufacturing for the de-
velopment of 3D porous scaffolds. In: Biomaterials for Bone Regeneration. Elsevier,
printing temperature 105–145 °C potentially extends the applicability pp. 149–201.
of this technology to a wider range of active pharmaceutical in- Husken, D., Gaymans, R., 2009. The tensile properties of poly (ethylene oxide)-based
gredients. A novel radiator-like paralleled plate geometry oral doses segmented block copolymers in the dry and wet state. J. Mater. Sci. 44 (10),
2656–2664.
containing widely used biodegradable polymer species (PEOs and PEG) Kennedy, M., Peacock, A., Mandelkern, L., 1994. Tensile properties of crystalline poly-
was reported. By using this architecture, it was possible to accelerate mers: linear polyethylene. Macromolecules 27 (19), 5297–5310.
drug release and overcome polymer hindrance of theophylline release Kim, C.J., 1998. Effects of drug solubility, drug loading, and polymer molecular weight on
drug release from Polyox tablets. Drug Dev. Ind. Pharm. 24 (7), 645–651.
through PEO swelling and erosion. These findings are essential in the Kojima, H., et al., 2008. Extended release of a large amount of highly water-soluble
development of next-generation personalised drug delivery doses using diltiazem hydrochloride by utilizing counter polymer in polyethylene oxides (PEO)/
specialised polymer/polymer blends purposely optimised for FDM 3D polyethylene glycol (PEG) matrix tablets. Eur. J. Pharm. Biopharm. 70 (2), 556–562.
Li, Q., et al., 2018. Preparation and investigation of novel gastro-floating tablets with 3D
printing. extrusion-based printing. Int. J. Pharm. 535 (1–2), 325–332.
Maggi, L., Bruni, R., Conte, U., 2000. High molecular weight polyethylene oxides (PEOs)
Conflict of interest as an alternative to HPMC in controlled release dosage forms. Int. J. Pharm. 195
(1–2), 229–238.
Moroni, A., Ghebresellassie, I., 1995. Application of poly(oxyethylene) homopolymers in
None declared. sustained-release solid formulations. Drug Dev. Ind. Pharm. 21, 1411–1428.
Nasereddin, J.M., et al., 2018. Development of a Simple Mechanical Screening Method for
Appendix A. Supplementary data Predicting the Feedability of a Pharmaceutical FDM 3D Printing Filament. Pharm.
Res. 35 (8), 151.
Palo, M., et al., 2017. 3D printed drug delivery devices: perspectives and technical
Supplementary data to this article can be found online at https:// challenges. Expert Rev. Med. Devices 14 (9), 685–696.
doi.org/10.1016/j.ijpharm.2019.04.016. Pereira, B.C., Isreb, A., Forbes, R.T., Dores, F., Habashy, R., Petit, J.B., Alhnan, M.A., Oga,
E.F., 2019. ‘Temporary Plasticiser’: A novel solution to fabricate 3D printed patient-
centred cardiovascular ‘Polypill' architectures. Eur. J. Pharm. Biopharm. 135,
References 94–103.
Pietrzak, K., Isreb, A., Alhnan, M.A., 2015. A flexible-dose dispenser for immediate and
extended release 3D printed tablets. Eur. J. Pharm. Biopharm. 96, 380–387.
Alhijjaj, M., Belton, P., Qi, S., 2016. An investigation into the use of polymer blends to
Prasad, L.K., Smyth, H., 2016. 3D Printing technologies for drug delivery: a review. Drug
improve the printability of and regulate drug release from pharmaceutical solid
Dev. Ind. Pharm. 42 (7), 1019–1031.
dispersions prepared via fused deposition modeling (FDM) 3D printing. Eur. J.
Sadia, M., et al., 2018. From ‘fixed dose combinations’ to ‘a dynamic dose combiner’: 3D
Pharm. Biopharm. 108, 111–125.
printed bi-layer antihypertensive tablets. Eur. J. Pharm. Sci. 123, 484–494.
Alhnan, M.A., et al., 2016. Emergence of 3D printed dosage forms: opportunities and
Shojaee, S., et al., 2015. An Investigation on the effect of polyethylene oxide con-
challenges. Pharm. Res. 33 (8), 1817–1832.
centration and particle size in modulating theophylline release from tablet matrices.
Apicella, A., et al., 1993. Poly(ethylene oxide) (PEO) and different molecular weight PEO
AAPS PharmSciTech 16 (6), 1281–1289.
blends monolithic devices for drug release. Biomaterials 14 (2), 83–90.
Skowyra, J., Pietrzak, K., Alhnan, M.A., 2015. Fabrication of extended-release patient-
Arafat, B., et al., 2018. Tailored on demand anti-coagulant dosing: an in vitro and in vivo
tailored prednisolone tablets via fused deposition modelling (FDM) 3D printing. Eur.
evaluation of 3D printed purpose-designed oral dosage forms. Eur. J. Pharm.
J. Pharm. Sci. 68, 11–17.
Biopharm.
Tagami, T., et al., 2017. 3D Printing factors important for the fabrication of poly-
Awad, A., et al., 2018. 3D printed medicines: a new branch of digital healthcare. Int. J.
vinylalcohol filament-based tablets. Biol. Pharm. Bull. 40 (3), 357–364.
Pharm. 548 (1), 586–596.
Wissbrun, K.F., 1980. Observations on the melt rheology of thermotropic aromatic
Beech, D.R., Booth, C., 1970. Thermodynamic melting point of poly(ethylene oxide). J.
polyesters. Br. Polym. J. 12 (4), 163–169.
Polym. Sci., Part C: Polym. Lett. 8 (10), 731–734.
Zhang, F., McGinity, J.W., 1999. Properties of sustained-release tablets prepared by hot-
Bliley, J.M., Marra, K.G., 2015. Polymeric biomaterials as tissue scaffolds. In: Stem Cell
melt extrusion. Pharm. Dev. Technol. 4, 241–250.
Biology and Tissue Engineering in Dental Sciences. Elsevier, pp. 149–161.

105