Polymer Testing 108 (2022) 107499

Contents lists available at ScienceDirect

Polymer Testing
journal homepage: www.elsevier.com/locate/polytest

3D printing and post-curing optimization of photopolymerized structures:

Basic concepts and effective tools for improved
thermomechanical properties
Martina Štaffová a, František Ondreáš a, b, Juraj Svatík a, Marek Zbončák c, Josef Jančář a,
Petr Lepcio a, *
a
Central European Institute of Technology, Brno University of Technology, Purkyňova 123a, 612 00, Brno, Czech Republic
b
Contipro a.s., Dolni Dobrouc 401, 56102, Dolni Dobrouc, Czech Republic
c
Prusa Polymers a.s., Partyzánska 188/7a, 170 00, Praha, Czech Republic

A R T I C L E I N F O A B S T R A C T

Keywords: The final thermo-mechanical properties of structural parts fabricated by masked stereolithography (MSLA) are
3D printing highly determined not only by the processing parameters, but also by the post-processing methods. Improper
Masked stereolithography implementation of post-treatment often leads to underperforming printouts. A novel tool for complex charac­
Post processing
terization of 3D printed bodies was developed and systematically demonstrated on a commercial free-radical
DMA
photopolymerization (FRP) resin. The method relies on superimposed static and oscillatory mechanical test
Print orientation
Network density combining the heat deflection temperature (HDT) measurement together with the dynamic mechanical analysis
(DMA) in a single test for fast and reliable characterization of parameters determining the curing behaviour of
the photopolymer. The influence of post-curing time was addressed with a special focus on network density.
Furthermore, the print orientation, having a high impact on mechanical properties, is discussed with a particular
regard on the residual stress mitigation in future applications, such as 3D-printed cellular bodies.

1. Introduction Recent developments have reduced the cost of 3D printers, enabling

desktop fabrication of 3D objects even at home, thereby expanding its
3D printing, also known as additive manufacturing (AM) has expe­ application field from science and industry to small businesses and
rienced a rapid growth in recent decades. Freedom of design, custom­ general public. Among the 3D techniques, stereolithography (SLA) is the
ization, waste reduction, fast prototyping, high precision, avoiding the earliest method patented in 1986 by Hull [6] and also the first
use of tools and ability to manufacture complex structures are the main commercially employed one. It has been reported that almost 50% of the
benefits of 3D printing [1]. The main disadvantage is the heterogeneity 3D printing market can be attributed to photopolymers [7]. Photo­
of the 3D printed bodies due to the defects between the adjacent print polymers have been used in fields such as lightweight engineering [8],
layers or voxels or inhomogeneous conversion of the monomer ultrasonic sensing [9] or functional constructs [10]. While they have yet
throughout the printouts. In general, the parts 3D printed by various to be approved for long-term surgical implantation, they have been used
techniques often fall behind the bulk processed materials in mechanical in medical device applications [11] such as hearing aids, microneedles
properties unless the print layers/voxels are properly fused together [2, [12] or drug-loaded scaffolds with controlled release properties [13].
3]. Furthermore, the heterogeneous character makes the overall me­ However, there are still numerous issues that need to be addressed in
chanical performance sensitive to the print orientation which de­ order to full scale the SLA 3D printing. Moreover, due to the layered
termines the distribution of defects [4]. On the other hand, smart manufacturing process, the products tend to be anisotropic, with the
designs could engineer weak spots on purpose to modify the overall boundary between layers representing weak regions with maximum
mechanical performance. Even though the ultimate strength is reduced, residual stresses [14].
a local failure at these weak spots could prevent a catastrophic fracture The traditional SLA technique employs an irradiation with a
in brittle materials [5]. controlled ultraviolet (UV) spotlight or laser to polymerize the liquid

* Corresponding author. Postal address: CEITEC VUT, Purkynova 123a, 612 00, Brno, Czech Republic.
E-mail addresses: [email protected] (M. Štaffová), [email protected] (P. Lepcio).

https://doi.org/10.1016/j.polymertesting.2022.107499
Received 22 December 2021; Received in revised form 20 January 2022; Accepted 2 February 2022
Available online 4 February 2022
0142-9418/© 2022 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
M. Štaffová et al. Polymer Testing 108 (2022) 107499

resin point by point, which results in a patterned layer of the crosslinked technique for engineers and scientists as well as for the rapidly growing
polymer [15]. Masked stereolithography (MSLA) was derived from the small business, and general public community.
SLA technique by putting a mask with a controlled shape between the
light source and the resin tank to cure a specific 2D pattern. These 2. Material and methods
methods exhibit higher build speed, excellent surface quality and pre­
cision with a typical feature resolution of tens to hundreds of micro­ 2.1. Materials
meters depending on the technological parameters (light source,
wavelength, pixel size, material, etc.). The typical curing behaviour of A commercial UV curable acrylate-based transparent clear resin (CR)
common free radical polymerization resins is characterized by a high obtained from Shenzhen Yongchanghe Technology Co., Ltd. was used as
built speed of a polymer network (high-reactivity), fast transition from obtained. According to its UV-VIS spectra recorded by Jasco V-730 UV-
liquid to solid (gelation) and a quick setting of the final mechanical VIS spectrophotometer and compared to the neat TPO (RAHN,
properties (curing), which leads to increased stiffness as a result of Switzerland) dissolved in acetonitrile (Fig. S1), the resin contained 6%
changes on molecular level of the photopolymer material [16,17]. A of the diphenyl (2,4,6-trimethylbenzoyl)-phosphine oxide (TPO) free
crosslinking of the system with higher functional acrylate oligomers and radical photoinitiator. This resin was selected as a representative of
monomers is necessary in order to increase the network density and thus many common commercial photocurable resins, either for non-
the thermomechanical stability suitable for the 3D printing process [18]. demanding applications or for prototyping, which have similar chem­
Numerous research activities on kinetics and characterization of istry, penetration depth and critical energy. Further thermo-mechanical
photopolymers have been performed [19–22]. Hong et al. conducted properties of the material are assessed below.
research on mechanical properties of UV-curable 3D printing materials,
along with their anisotropic and size dependent behaviour [23]. The 2.2. MSLA 3D printing
anisotropic properties were clearly seen in tensile toughness and elon­
gation at break with respect to a printing direction, while it was not Samples were 3D printed with Prusa SL1 printer (Prusa Research,
observed in their Young’s modulus and tensile strength. Kotlinski per­ Czech Republic), based on the Mask Stereolithography technology
formed an in-depth analysis of commercial AM materials with respect to (MSLA). The irradiation intensity of the 3D printer was 0.661 mW cm− 2,
their mechanical properties, including guideline for design engineers measured at the surface of the FEP foil in the vat with the Flame-S-UV-
[24]. The diffusion-controlled kinetics of curable 3D printing materials VIS-ES spectrometer (Ocean Insight, USA). The maximum light intensity
has been studied by Kim et al. [25] and kinetic model was developed to was found at 405 nm (Fig. S1). Standard FEP foil with thickness of 150
describe the limited conversion of cure as a function of isothermal curing μm was used. The exposure time was selected according to the Jacobs
temperatures. Gao et al. [26] presented photopolymerization kinetic working curves as described further in text. All specimens were printed
model to predict printing effects, while it was concluded that printing with the exposure time of the initial layer of 15 s. Further layers were
settings had a great influence on printing quality for paradoxical rela­ printed with the exposure time of 10 s. The layer thickness was set to 50
tionship between print speed and resolution. By Hofstetter et al. the cure μm. If not specified otherwise, the samples were printed in x direction, i.
degree of a cured layer was evaluated in order to find a way to char­ e., with the narrower face aligned to the built platform plane. Printed
acterize photopolymers [16] concluding that with the combination of specimens were washed in isopropanol for the shortest possible time
Jacobs working curve (cure depth), the mechanical properties and (less than 15 s) to avoid monomer extraction, swelling, and the conse­
double bond conversion (DBC) from the photorheometer (cure degree), quent deterioration of the mechanical properties [27,30,31]. It was
it is possible to completely characterize the curing behaviour of photo­ observed that the exposure time of the fresh printout to the cleaning in
polymers. Further the great effect of post-curing and solvent isopropanol had a significant influence on sample appearance and the
post-processing methods on mechanical properties have been studied presence of defects. A significant effect on thermomechanical properties
[27,28]. However, there is still need for better understanding of the and final application performance can be expected. Thus, all samples
photopolymerization process as well as for utilized characterization were cleaned for less than 15 s in this study. Nevertheless, the effect of
methods of novel resins. solvent washing on the post-treatment was beyond the scope of the
Herein, we present a novel approach to characterize the thermo­ current study even though it is of great importance. Post processing was
mechanical properties, which shall deliver a powerful tool for future performed with the Prusa CW1 curing and washing unit with approx.
excessive studies of photocurable 3D printing resins and their use in irradiation intensity of 10 mW cm− 2. The samples were dried and cured
engineering applications. We present an effective, reliable, time saving, at 35 ◦ C. The influence of sample thickness, the curing time, print
and relatively simple hybrid method combining the DMA and HDT orientation and position of the sample on the printing stage on ther­
measurements into a single test. The advantages in comparison to con­ momechanical performance was studied. All specimens were printed
ventional methods of HDT and DMA measurements methods are lower off-centered, to provide the most uniform light distribution on the
material consumption, time, effort, and capacity savings due to faster printing plate.
sample preparation and measurement, and collection of additional data
in a single run compared to situation where only one test is carried out. 2.3. Thermomechanical properties analysis
Furthermore, it lowers the equipment requirements, if only one instru­
ment is needed instead of two. It originated from the fundamental A method for simultaneous DMA and HDT analysis was performed
investigation of polymer glass dynamics via superimposed small static with RSA G2 from TA Instruments to establish material constants and
stress typical for HDT measurement and oscillatory deformation printing parameters. The stiffness measured through the dynamic
employed by the DMA [29]. The thermomechanical properties were oscillatory deformation represented by storage modulus G’ of the stud­
studied with a representative commercial free radical polymerization ied materials was evaluated at different temperatures and the glass
(FRP) photo-resin. The parameters monitored in this study were the transition temperature Tg was determined from the maximum of the loss
curing time, specimen thickness and orientation on the print platform. factor (tan δ). Furthermore, HDT measurement was based on the ASTM
The presented hybrid method showed a great sensitivity to all these D 648-07 Standard Test Method for Deflection Temperature of Plastics
parameters and correlated them to the network density, a key structural Under Flexural Load in the Edgewise position. The size of the test
parameter established by this technique which directly depends not only specimens was proportionally reduced against the standard body to
on the resin composition and the photochemical reaction upon the 3D 48.5 × 5 x 2–5 mm3. The specimens were loaded in three-point bending
printing, but also on the subsequent processing steps. The results will setup with 40 mm distance between supports in the edgewise direction.
help improve the understanding of the photopolymerization and MSLA HDT was determined at a deformation that corresponded to a deflection

2
M. Štaffová et al. Polymer Testing 108 (2022) 107499

of 0.25 mm of the normalized body, which equals to 0.195% flexural All samples were printed as squares with the dimension of (10 × 10)
strain for the samples used. The flexural strain was calculated according mm2, while the curing time varied between 10 and 60 s. After printing,
to equation (1): the squares were cleaned with isopropanol and their thickness was
measured (minimum five times per each sample).
600⋅s⋅h
εf = (1)
L2
2.5. FTIR measurement of double bond conversion
Where s is the deflection, h is thickness and L is the span. The frequency
of oscillating deformation was set to 1 Hz and the strain amplitude was In order to find out the conversion ratio variation throughout the
0.002%. The imposed stress of 0.455 MPa was applied by the upper sample’s thickness, the FTIR measurement was taken with VERTEX 70v
geometry and the heating from 30 to 160 ◦ C at the rate of 2 ◦ C/min was FTIR equipped with Hyperion 3000 Microscope in ATR mode. The
mediated by hot air. The validity of this presented method is discussed in specimen was printed with dimensions of 48.5 × 5 × 5 mm3 and post-
more detail in the results section. Moreover, the results were correlated cured for 1 min with the Original Prusa CW1 curing machine. The
to the hardness which was measured with a handheld Shore D durom­ sample was cryofractured in liquid nitrogen to expose the bulk body and
eter (SMT, Netherlands) with a 30◦ cone, radius of 0.1 mm and the measured at eleven evenly distributed spots along the thickness of the
maximum force of 44.5 N, a fast, cheap, and simple technique which is broken plane.
available not only to specialized laboratories, but could be easily
adopted also by small business and hobby users. 2.6. Structural observation

2.4. Photopolymerization assessment The surface structure of the printed specimens was investigated with
the Confocal Laser Scanning Microscope Olympus Lext OLS4100
The photopolymerization reaction was assessed by photo-DSC and equipped with a 405 nm laser. The SEM observation was performed on
photo-rheology using DSC Discovery and ARES G2, respectively, both cryo-fractured samples coated with a 10 nm sputtered gold (ACE 600,
from TA Instruments. The uncured resin loaded in the instruments was Leica) with a Mira 3 XMU microscope (Tescan, Czech Republic) in low-
illuminated by a constant 405 nm LED light source with a power in­ vacuum mode at 10 kV acceleration voltage using the LVSTD secondary
tensity of 0.661 mW cm− 2 measured inside the instrument at the sample electron detector designed for low vacuum operation and observation of
location, i.e., matching the power density of the SL1 printer. The onset of organic low conductive samples.
illumination was set to the time of 0 s. Photo-DSC was performed in an
open aluminium pan with 15.0 μl of sample at a constant temperature of 3. Results and discussion
25 ◦ C against an empty reference pan which was also illuminated by the
same light. The results were compared to high intensity curing using the The chemical composition, functionality, interactions, stiffness, and
same photo-calorimetric setup and the OmniCure S2000 curing unit chain length of the monomers strongly affect the photopolymerization
equipped with a 200 W high pressure mercury vapor UV lamp with a reaction, safety, and other properties of the final systems. Thus, it is of
power density of 30 W cm− 2. Photo-rheology was conducted under great importance to have a reliable testing method for characterizing
ambient conditions in steady oscillation at frequency of 4 Hz, strain these materials. Thermomechanical properties directly reflect the
amplitude of 2%, and gap of 100 μm. A special disposable UV-curing structure of the analysed system. Therefore, they represent a suitable
plate-plate geometry (transparent acrylic upper geometry, aluminium tool for evaluating the influence of the processing and post-processing
lower geometry) with a diameter of 25 mm was used to illuminate the parameters on the photopolymerized 3D printed systems.
top surface of the sample. The network density (υe) could be directly determined by dynamic
The samples for the cure depth (Cd) measurement were printed using mechanic analysis (DMA) through measuring the modulus at a tem­
the SL1 printer (Prusa Research, Czech Republic). The resin was poured perature well above the Tg, in the rubbery region. The network density νe
into the printer’s vat and a single layer square shape was photo­ is estimated from the rubbery modulus using the theory of rubber
polymerized in the absence of the printing platform. The cure depth of elasticity (Eq. (5)) [33,34]:
the layer was measured with a micrometre as a function of the exposure
G0N
energy by varying the exposure time. Jacobs working curve was created νe = 2(1 + υ) , (5)
as a linear-logarithmic plot of cure depth dependence on the exposure RT
energy (Emax). The intersection of the working curve with the x-axis where υ is the Poisson’s ratio, R is the gas constant, T is the temperature,
gives the critical exposure Ec, e.g., the exposure at which the resin so­ and G0N is the rubber-plateau modulus. The network densities were
lidification starts to occur. The slope of logarithmic-linear plot repre­
calculated assuming the material is incompressible (υ = ½), using the
sents the penetration depth (Dp) of light into a resin at which the
rubber-plateau modulus G0N determined at the fixed temperature of
irradiation dose Ez equals to the Emax/e where e is the Euler’s number (see
135 ◦ C. Furthermore, the stiffness represented by storage modulus E′ of
Eq. (2)). Through the application of the Beer-Lambert law, the theo­
the studied materials could be evaluated at different temperatures, while
retical relation between the resin characteristics and exposure can be
the glass transition temperature Tg, at which the segmental motion of
developed (see Eq. (3)) [32].
polymer chains is released and the stiffness decreases significantly, was
( )
Emax determined from the maximum of loss factor (tan δ).
Cd = Dp ⋅ln = Dp ⋅ln Emax − Dp ⋅ln Ec (2)
Ec Another common tool for evaluating the thermomechanical prop­
erties is the maximum deformation (Dmax) and the heat deflection
( )
z temperature (HDT) which indicates the temperature at which the body
Ez = Emax ⋅exp − (3)
Dp bends at a given deflection under the constant load. These parameters
are also directly related to the structure of the analysed system. We have
When Eq. (2) is fitted with a simple logarithmic function y = a ln x +
adopted the HDT measurement based on the ASTM D 648-07 Standard
b, where the x and y represent the exposure energy and the cure depth,
Test Method for Deflection Temperature of Plastics Under Flexural Load
respectively, the penetration depth is directly equal to the parameter a
in the Edgewise position, proportionally reducing the size specified for
while the critical exposure could be evaluated as (Eq. (4)):
the normalized specimens to 48.5 × 5 x 2–5 mm3. The static stress was
Ec = e−
b
a (4) maintained at 0.455 MPa.
In order to save time and material, a combined DMA and HDT

3
M. Štaffová et al. Polymer Testing 108 (2022) 107499

measurement was developed which was performed on RSA G2 (TA In­

struments, USA). This method, originating from the fundamental
investigation of the polymer glass dynamics [29], superimposes the
small static stress used by the HDT measurement with the oscillatory
stress of the DMA method to establish the material constants and pro­
cessing parameters in a single run (Fig. 1). The hybrid method is
particularly convenient for characterizing 3D printed materials because
of the rather long processing times and often the limited build volume.
Nevertheless, this method is only eligible to test solid specimens. Thus,
firstly are presented the results on photocuring assessment obtained by
other techniques which were used to obtain the basic printing parame­
ters. Further continue the discussion of the mechanical properties and
network density.
The glassy region at low temperatures was characterized with a
relatively weak dependence of the storage modulus on the temperature
and low tan δ values (Fig. 1). As the heating continued, a drop in storage
modulus and peak of tan δ appeared for all samples, marking the glass
transition region where the material losses its stiffness. The rubbery
behaviour with its characteristic rubbery plateau of storage modulus
and low tan δ values followed at even higher temperatures. The terminal Fig. 2. Jacobs working curve for the tested resin. Standard deviation is below
zone and macroscopic flow of the samples was not observed due to their 8 μm for all samples and therefor is not visible in the graph.
highly crosslinked structure. However, failure of the samples at specific
temperature associated with the printing direction was observed which illuminated with low intensity (0.661 mW cm− 2) blue LED light (405
is described further in the text. nm) with the maximum conversion rate reached at (12.3 ± 0.3)
mJ⋅cm− 2 (~18.6 s) followed by another local maximum at (213.6 ±
3.1. Photocuring assessment 19.5) mJ⋅cm− 2 (~323.1 s) (Fig. 3 left). The second peak is reportedly
connected to the reaction of the heavier monomer (or oligomer) in a
The parameters of the photocuring process were evaluated using the mixture containing at least two monomers with different molecular
Jacobs working curves, photo-DSC and photo-rheology to properly weight [36]. We note that the second peak is not commonly observed in
adjust the printing parameters. The Jacobs working curve of the tested photo-DSC results when a high intensity (30 W cm− 2) wide-range UV
resin is shown in Fig. 2. The critical exposure (Ec) and penetration depth spectrum mercury lamp is used for the curing (Fig. 3 right). Interest­
(Dp) were calculated according to equations (3) and (4). The resin ingly, the threshold energy of 9.6 mJ cm− 2 at which the cure depth
showed a typical curing behaviour for acrylate photopolymers, which started to grow faster (Fig. 2) slightly preceded the maximum of the first
can be characterized by high reactivity and fast gelation and curing. The peak (Fig. 3 left) but it is obvious that much higher exposure doses are
penetration depth must be larger than the printing layer thickness in required to fully cure the resin. The total curing enthalpy was estab­
order to ensure sufficient adhesion between the layers and avoid lished to (− 343.8 ± 4.1) J⋅g− 1. We point out that the maximum con­
delamination [35]. Furthermore, the Jacobs working curve revealed a version rate took place at varying time depending on the volume of the
sudden increase in the cure depth rate at about 9.6 mJ cm− 2 (~14.5 s). cured sample (data shown for 15.0 μl). In the contrary, the total curing
The data beyond this critical point could be fitted with an independent enthalpy was quite insensitive to this parameter but that would not be
logarithmic fit which yielded higher Dp as well as the Ec compared to the expected unless for very thick samples.
initial photopolymerization reaction. The process behind this sudden The photo-rheology (Fig. 4) revealed that the gel point evaluated at
change in the curing behaviour could be better understood from the the crossover of storage G′ and loss G” moduli took place at 3.0 mJ cm− 2
photo-DSC and photo-rheological results which discussion follows. (~4.6 s), corresponding well to the critical energy of 3.3 mJ cm− 2
The photo-DSC revealed two exothermic peaks when the resin was established by the Jacobs working curves. At higher exposure doses, the
increase of G′ has saturated, accompanied with a slight overshoot at G”
and a peak at the tan δ at 7.1 mJ cm− 2 (~10.8 s). This may possibly
relate to the vitrification point, i.e., the moment when the system turned
into glassy state. That happens when the gradually increasing Tg of the
cured system exceeds the experimental temperature at which the
curing/measurement takes place. The detailed evaluation of Tg in the
printed bodies will be discussed later. Interestingly, the sudden drop in
the curing heat flow at 12.3 mJ cm− 2 (Fig. 4 left) took place 7.8 s after
the vitrification point (Fig. 4 right). The transition to glass region would
be expected to slow down the diffusion of species within the sample and
hinder the propagation of the ongoing reaction, but this phenomenon
may also possibly relate to the full consumption of the lighter monomer.
The curing energy of 5.7 mJ cm− 2 (~8.6 s) required to polymerize a 100
μm layer (Fig. 2), i.e., to span across the gap used for the rheological
measurements, is lower than the energy required for the vitrification to
occur. Instead, it correlates with the exposure dose at which the rapid
growth of G’ and G” is replaced by nearly indistinct change in properties
(Fig. 4 right). We note that the plate-plate geometry setup of the photo-
rheometer may struggle to transfer the stress into very stiff samples,
rendering the procedure insensitive to further structural changes beyond
Fig. 1. Graph showing main values read from the developed technique this point.
combining DMA and HDT measurement of photo-resins. (color in print)

4
M. Štaffová et al. Polymer Testing 108 (2022) 107499

Fig. 3. Normalized heat flow recorded by a photo-DSC as a function of exposure time and exposure energy upon irradiation by the low intensity 405 nm UV LEDs
(left) and high intensity wide-spectrum mercury lamp (right).

Fig. 4. Photorheological assessment of the UV curing process – dependence of storage (black) and loss (red) moduli and loss factor tan δ (blue) on the exposure time
and exposure energy (left) with the detail of the gel point and vitrification point (right). (For interpretation of the references to colour in this figure legend, the reader
is referred to the Web version of this article.)

3.2. The effect of printing orientation observed to highly depend on the printing orientation (Fig. 6). Printing
in the x direction showed the highest resistance to failure due to the non-
The Jacobs working curves, photo-DSC and photo-rheology served as isothermal creep thanks to the most effective layer alignment in the
a basis to establish the most convenient exposure time of the MSLA 3D given load direction. The force was applied in the direction normal to
printing experiments, which was set to 15 s (~9.9 mJ cm− 2) for the first the printed layers; bending the whole layers when applied. A growing
layer and to 10 s (~6.6 mJ cm− 2) for the subsequent layers. Neverthe­ crack had to pass through several boundaries formed by layered struc­
less, these results gave no clue on the molecular structure and other ture of weak and compact areas before resulting into the ultimate failure
important parameters such as the printing orientation or the network of the specimen. On the other hand, specimens printed in the z-direction
density which strongly influences the mechanical performance [16]. showed lowest thermomechanical resistance to the propagating crack
Two sets of specimens with various curing times were printed in three during the deformation at higher temperatures because the weak layer
different directions x, y, z to investigate the influence of printing boundaries were aligned with the applied stress. Thus, the adjacent
orientation on thermomechanical performance (Fig. 5). The first set was layers were separated by a crack which started at the tensile-deformed
measured without post-curing (“green body”) while the second one was side of the sample and grew easily through the weak layer boundary.
post-cured for 30 min (~ 18.0 J cm− 2). A pronounced effect of the The printing in the y-direction exploited intermediate thermomechan­
printing layers’ orientation relative to the applied force was observed. ical resistance when compared to x and z direction (Fig. 6). The possible
The classic HDT measurement as well as our modified method applies a explanation might be the fact that the propagating crack does not need
constant load to the sample. In addition, a non-isothermal creep to cross the interlayer region where it can branch and delaminate the
occurred at higher temperature, causing a mechanical failure of the neighbouring layers, thus, dissipating more energy.
specimen at a certain point [29]. The deformation and temperature at The photos of uncured sample printed in different directions after the
break could be used as material characteristics for the specified heating mechanical stress are shown in (Fig. 7). The distance between layers was
protocol. The latter effectively replaces the time to break which is usu­ measured to be approx. 50 μm as expected according to the set layer
ally evaluated in isothermal creep tests. thickness of 50 μm. The same layer distribution was observed for all
Significant differences in samples’ temperature resilience were uncured samples printed in x, y and z direction. Uncured samples

5
M. Štaffová et al. Polymer Testing 108 (2022) 107499

Fig. 5. Schematic representation of the print orientation as the specimens were positioned on the printing plate (left) and in the combined DMA-HDT test with
corresponding layer orientation relative to the applied force (right).

Fig. 6. The influence of print orientation on temperature at break and Tg. The
sample with dimensions 48.5 × 5 × 3.5 mm3 printed in the x direction and
cured for 30 min did not break until the end of the test. (color in print)

showed the highest concentration of cracks combined with worse ther­

momechanical performance, i.e., low T at break, polymer network Fig. 7. Uncured samples printed in x, y, and z direction after DMA measure­
density, HDT, and Tg, caused by the lower interconnection of individual ment, revealing print-direction dependent delamination (top); SEM image of an
uncured sample showing printed layers as was schematically presented in Fig. 5
layers and decreased double bond conversion. However, after 30 min
with dimensions of approx. 50 μm (bottom left) and a sample after exposure to
post curing the boundaries between the individual layers were not
mechanical stress with cracks passing through the boundaries between the
distinguishable anymore. This correlates well with the increased poly­ layers (bottom right).
mer network density and the improved mechanical properties of the
material, which is further discussed. An uncured sample after exposure
The failure temperature and the mechanical resistance increased
to mechanical stress is shown in Fig. 7 bottom right. The cracks passed
significantly after post-curing for all printing orientations due to the
through the boundaries between the layers of the sample, as expected.
stronger interconnection of individual layers because of the increased
This phenomenon was not observed for none of the cured samples at any
double bond conversion. These results indicated pronounced improve­
print direction.
ment of the non-isothermal creep resistance which decreased in the

6
M. Štaffová et al. Polymer Testing 108 (2022) 107499

order x > y > z printing direction. Hence, it would significantly extend post-curing times and exposure doses. The storage modulus is a measure
the service lifetime of these materials in applications with demands on of sample’s elastic behaviour associated with its stiffness. We have
good mechanical performance, especially at higher temperatures. The selected the representative temperature of 35 ◦ C for the precise com­
glass transition temperature remained the same for all directions as this parison of the glassy modulus between the samples. This point lay deep
property was not influenced by the load direction (Fig. 5) and solely in the glassy region of all the tested specimens while the heating cycle
depended on the molecular structure within the material regardless the has already stabilized after starting the heating procedure at 30 ◦ C. The
layer orientation and stress transfer between the neighbouring layers or tan δ can be interpreted as a damping factor reflecting the energy
individual voxels. The printing orientation had also a pronounced dissipation of the sample at the given temperature. Moreover, the values
impact on the temperature at break which differed for the uncured of G0N , υe, Mc, and Tg are also stated in Supplementary data (Table S1).
samples. A longer post-curing reduced the difference, as the layers’ The prolonged curing time and higher exposure doses led to the
interconnection was improved, but it was still significant. significant shift of the Tg and the whole glass transition region towards
The effect of the print direction was also significant for the deter­ higher temperatures (Fig. 10). There was also a noticeable increase in G0N
mination of HDT and Shore D hardness (Fig. 8). A noticeable difference at 35 ◦ C which was partially connected to the increased network density,
was observed in print directions x and y even at low curing times. The as will be explained later, but also to the higher relative distance from
HDT values for x-direction showed better results. The difference in the glass transition due to the increased Tg.
various directions gradually decreased as the material became more and Further information was extracted from the temperature dependence
more resistant to deformation. Thus, it is particularly important to of the flexural deformation induced by the superimposed static stress
consider the loading direction regarding the printing direction for the (Fig. S2). After a short period dominated by the sample’s volume
correct evaluation of thermomechanical properties and HDT. It can be expansion, manifested by negative strain values, the deformation started
concluded that the combined DMA-HDT analysis can conveniently to increase, i.e., bending the testing specimen. That corresponded to the
visualize the mechanical properties varying with the printing layers’ increased segmental motion of the polymer network and softening of the
orientation respective to the applied force and stress transfer of applied material incurred at the glass transition region. The maximum of the
force within the material, while the thermal properties were insensitive detected deformation shifted to lower values and higher temperatures
to the orientation. for the post-cured samples due to the increased network density
(Fig. S2). The deformation maximum was followed by the rubber elastic
region (Fig. S2).
3.3. The effect of post photo-curing time
The polymer network density, evaluated according to eq. (5),
increased with the prolonged post-curing time, as expected (Fig. 10).
Both professionals and hobby users often purposely decrease the
Subsequently, higher exposure doses also led to the increase in all the
post-curing time of the printed object to shorten the production time.
investigated thermomechanical properties (Table S1) and hardness. The
Moreover, they commonly neglect the requirements for the irradiation
increase in glass transition temperature, HDT and network density was
of the printed objects from all sides of the printout. These flaws might
asymptotic, i.e., tending to a limit (Fig. 10). The limit value is supposed
have serious consequences for the part durability which can be critical if
to correspond either to the completely cured system or, more precisely,
we consider engineering of load bearing applications or applications
to a fully cured system under the given post-processing conditions
requiring low or no migration of molecules from and/or into the object.
determined among others by the light wavelength, exposure dose, and
It may also prove critical for various structures with hidden surfaces
temperature in the curing chamber.
such as the inner walls of cells and cavities which might be inaccessible
Uncured 3D printing resins contain large free macroradicals unable
to the post-curing light.
to react at common temperatures beyond the vitrification point due to
The curing time and the total exposure energy had a major influence
the hindered diffusion, which results into imperfect UV curing with a
on the mechanical performance. The temperature dependence of the
lower conversion. Some materials with low double bond conversion
printed objects’ storage modulus and tan δ is shown in Fig. 9 for different
might show problems during the post-curing at higher temperatures,
especially when not fully cured, causing the emergence of cracks in the
objects (Fig. 7). This is given by inhomogeneous thermal expansion of
the object in individual layers where the conversion is a function of the
z-coordinate as well as by the non-homogenous distribution of the
temperatures at the surface and inner parts of the printed objects. In
combination with the extraction of free monomers from the object body
and diffusion of a thermally expansive washing solvent, such as iso­
propanol, it might lead to either formation of small cracks along the
layers and/or complete destruction of the printed object, deteriorating
the mechanical performance of the material [30]. The diffusion of the
free monomers or pending chain ends can be increased by heating the
system above its glass transition temperature, bringing about a full re­
action of the free radical. Even though the post-curing at elevated
temperature was not investigated in this study, a similar phenomenon
has been observed during the DMA-HDT measurements. The post cured
samples had yellowish colour caused by the light absorption of the free
radicals, while the samples after the measurement were colourless
(Fig. S3).
Two main trends were observed from the dependence of the ther­
momechanical properties on the curing time and exposure dose
(Fig. 10). Until 5 min curing time, there was an exponential increase in
Fig. 8. HDT and as a function of curing time for x and y print orientation all the studied properties. Further increase was rather linear,
combined with shore D hardness of the samples printed in x-direction showing approaching a limit value. Two characteristic curing times were estab­
similar trend. The negative influence of the layer orientation towards the lished from this behaviour - the minimal curing time (tmin) which may be
applied stress on HDT may be seen.

7
M. Štaffová et al. Polymer Testing 108 (2022) 107499

Fig. 9. DMA data for a) storage modulus and b) tan delta as a function of temperature for samples with the dimension of 48.5 × 5 × 3.5 mm3 cured for 0–60 min.
Curve shift caused by the increasing curing time is highlighted with an arrow pointing in the direction from 0 to 60 min.

suitable for basic undemanding applications as the initial sharp

improvement of all properties was already saturated, while the ideal
curing time (tideal) delivered the maximal thermomechanical properties
reachable at the given post-processing conditions. We note that these
values are connected to the layer thickness and may differ for thicker or
thinner layers. The tested resin reached the limit values after 60 min of
curing. This time was considered to be the ideal, since it showed the
maximal network density, plateau, deformation, HDT and storage
modulus at 35 ◦ C, i.e., the best thermomechanical properties obtained in
this study. The minimal post-curing time was set to 5 min (~3.0 J cm− 2).
We remind the reader that the storage modulus in the rubbery region
was the basis for the network density evaluation according to Eq. (5).
The network density (Fig. 11 left) increased with a similar trend as Tg
(Fig. 10). The molecular weight between entanglements decreased with
the increasing curing time as it is inversely proportional to the network
density. The polymer stiffness is dependent on the relative distance to
the glass transition temperature as well as on the network density [37].
Thus, the samples with the highest exposure doses were also the least
flexible. As far as the maximal deformation is concerned (Fig. 11 left),
the dependence on the curing time showed an inverse trend to the other
Fig. 10. Graph showing the thermomechanical properties evaluated by the studied parameters (Fig. 8). The temperature dependence on network
simultaneous DMA-HDT measurement as a function of post-curing time and the density for two main studied parameters, Tg and HDT in shown in Fig. 11
exposure energy E. right. It can be concluded that the network density is a driving structural
parameter for thermomechanical properties of photoresins, excluding

Fig. 11. Network density and maximal deformation as a function of post curing time for sample 48.5 × 5 × 3.5 mm3 (left) and the temperature dependence on
network density for two main studied parameters, Tg and HDT (right).

8
M. Štaffová et al. Polymer Testing 108 (2022) 107499

the influence of molecular structure of the monomer. experimental measurement error. This trend was further confirmed by a
The photopolymers used for SLA printing are usually described by microscopical FTIR-ATR measurement of double bond concentration
thermomechanical and mechanical properties such as tensile/flexural (see section 3.5). Thus, the UV light penetrated without significant loss
strength, elastic modulus, maximum elongation, HDT, Tg, or hardness, of energy through the sample with thickness between 2 and 5 mm. We
etc. [23,38–40] While some results found in the literature showed a note that the standardized testing specimen were 4 mm thick. This might
large impact of the printing setup on the mechanical performance [23], be caused by photobleaching of the TPO pohotoinitiator under the
other reported effects were far less pronounced [38]. Our results show irradiation with light during the post-curing. The light can thus pene­
that all the mechanical and thermomechanical parameters largely trate further to the inner parts of the 3D printed objects and propagate
depended on the printing setup as well as on the post processing steps the residual photoinitiator cleavage and curing reaction. However, it is
but the differences were variable, following the extent of curing and expected that for another photoinitiating system, larger specimen
correlating well to a single parameter – the crosslinking density. Previ­ thickness, lower exposure doses, lower light intensities, different light
ous studies have failed to quantify the crosslinking density as a function wavelengths with worse match between the curing light emission and
related to the specific combination of printing parameters, printed ma­ photoinitiator absorptivity, or resins filled with particles or
terial and cure dose, let alone correlating it to the material performance. light-absorbers, a different effect may be observed and negatively affect
This is particularly important when the slow reaction kinetics of the the crosslinking during post curing due to the light absorption.
oligomers in the mixed monomer/oligomer UV-curable formulations is
considered (Fig. 3) [36]. Thus, we propose to use this property measured 3.5. Double bond conversion
at different cure doses as a unification parameter for all UV-curable 3D
printing resins to offer a good predictability of mechanical and ther­ The curing conversion across the specimen thickness was determined
momechanical properties independent of the printing setup. We note using FTIR-ATR spectroscopical microscope (Fig. 13). The C– – C double
that this approach can also capture the changes related to the modifi­ bond absorption bands corresponding to acrylate groups can be seen at
cation of the resin, e.g. by adding a crosslinker [39]. 1639 and 1612 cm− 1 with a twisting vibration at 809 cm− 1. The single
carbon-carbon bonds (C–C) are visible at 2871 and 2938 cm− 1 and C–O
3.4. Specimen thickness bonds at 3443 cm− 1. The C– – O stretching vibration at 1724 cm− 1 was
used as an internal reference for the spectrum normalization. The rela­
Furthermore, the dependence of thermomechanical properties on tive concentration of the C–
– C and C–C bonds was then calculated as the

specimen thickness was studied to determine the in-depth post-curing area under the corresponding peak of the normalized spectra. The re­
effectivity. This parameter is important to many practical applications sults indicated an incomplete conversion of the C– – C which correlates

but becomes of an utmost importance in complex cellular structures with the further change in thermomechanical properties observed at
utilizing the cell-size and wall-thickness as tuning parameters of their longer exposure times (Fig. 10).
macroscale performance [5]. Samples with two dimensions fixed at 48.5 The C–– C of acrylate groups undergo a free radical polymerization

× 5 mm2 and a variable thickness of 5/4.5/4/3.5/3/2.5/2 mm were forming a cross-linked structure. The concentration of the C– – C double

printed from the tested resin and cured for 1 min. The curing time was bonds was studied along the depth of the specimen. “Candy-shell” effect
chosen in order to observe the largest possible difference; therefore, it can cause hardening of only a thin outer layer on the sample’s surface
was in the range where the properties changed very significantly with upon post curing due to exponential decay of light intensity along the
the curing time (Figs. 10 and 11). The DMA data showed overlapping depth of the specimen [27]. Based on this assumption, exponential
curves of storage modulus and tan δ as a function of temperature for decay of C–– C bonds was expected across the specimen. However, this

different specimen thickness (see Fig. S4). All studied parameters – the behaviour was not observed, for the specimen thickness between 2 and
deformation, glass transition temperature, molecular weight between 5 mm. The specimen was evenly cured throughout the body already
entanglements and network density (Fig. 12) remained unaffected by after 1 min of the post curing (Fig. 14) but that was likely connected to
sample thickness. Minor deviations were probably caused by the specific combination of parameters in the current experiment

Fig. 12. Network density as a function of specimen thickness for samples post- Fig. 13. Overlapped FTIR-ATR spectra of the printed resin post cured for 1 min
cured for 1 min. It can be seen, that thickness of the sample between 2 and 5 measured at ten spots evenly distributed across the cross-section of 5 × 5 mm.
mm have a negligible effect on network density and consequently on me­ The spectra confirm the conclusion, that the sample thickness between 2 and 5
chanical performance of the resin. mm did not affect the thermomechanical properties.

9
M. Štaffová et al. Polymer Testing 108 (2022) 107499

varied with the printing layers’ orientation respective to the applied

force and stress transfer of applied force within the material, while the
thermal properties were insensitive to the orientation. It was proven that
higher post-processing exposure doses increased the thermomechanical
properties, while having an asymptotic behaviour, i.e. tending to a limit.
Same trend, first the exponential increase followed by a slower linear
increase, was determined for all the studied material properties (HDT,
Tg, E and network density). Two curing times were selected from the
measured values as tmin suitable for undemanding applications while
keeping the process time at minimum and tideal with the maximum
network density and the best thermomechanical performance. The
method is expected to be applicable for all photopolymerized as well as
thermosetting systems. Furthermore, the thickness of the sample be­
tween 2 and 5 mm had in our case a negligible effect on the network
density, C– – C bonds conversion and consequently on the mechanical
performance of the printed structures. However, it is expected that
larger thickness, different curing times or presence of particles may
negatively affect the crosslinking and thermomechanical properties in
deeper layers due to uneven light penetration, especially when the post-
Fig. 14. Graf showing relative concentration of C– – C double bonds at 1564-
curing is applied. It was also observed that washing in isopropanol has
1668 cm− 1 across the 5 × 5 mm2 specimen with UV-thermal post curing for 1 large effect on thermomechanical properties due to the swelling
min, plotted with constant linear fit line, supplemented with photomicrograph resulting into delamination even though it was not systematically
of the measured sample with specific measured spots. studied. All these phenomena possibly relate to the presence of regions
at the layer boundaries which are not fully cured due to the hindered
allowing the even penetration of light through the specimen as propagation of the photoreaction beyond the vitrification point.
explained above. The negligible decrease in C– – C bond concentration Finally, we propose that the presented technique may represent a
was probably caused by the orientation arrangement in the CW1 curing new single-test standard helping to unify the available technical infor­
machine. mation on photopolymerized 3D printed bodies, in order to be able to
The surface of the printed specimens was studied with Confocal Laser fully predict the material characteristics. The acquired data considering
Scanning Microscope (CLSM). It was found that all specimens show the influence of preparation parameters on thermomechanical proper­
characteristic square pattern, which occurs before and also after the post ties of final 3D printed product may also significantly help in designing
treatment. The length of the square side pattern was measured and the processing conditions and thus enable proper choice of application
established to 0.044 mm which corresponded well to the pixels of the field.
SL1 printer LCD panel which was 0.047 mm per pixel. The fact that the
surface is structured may support the swelling of solvent during the post- Data availability statement
process cleaning. Despite this phenomenon was not systematically
investigated, it was observed that the exposure time of the fresh printout The processed data required to reproduce the above findings are
to the cleaning in isopropanol had a significant influence on sample available to download as an electronic supplementary file.
appearance and the presence of defects (Fig. S5). A significant effect on
thermomechanical properties and final application performance can be CRediT authorship contribution statement
expected. Thus, all samples were cleaned for less than 15 s in this study.
Nevertheless, the effect of solvent washing on the post-treatment was Martina Štaffová: Conceptualization, Data curation, Investigation,
beyond the scope of the current study even though it is of great Formal analysis, Writing – original draft, Writing – review & editing.
importance. František Ondreáš: Conceptualization, Formal analysis, Methodology,
Writing – review & editing. Juraj Svatík: Data curation, Formal anal­
4. Conclusion ysis, Investigation. Marek Zbončák: Conceptualization, Formal anal­
ysis, Validation, Writing – review & editing. Josef Jančář: Formal
A complex characterization method combining the measurement of analysis, Supervision. Petr Lepcio: Formal analysis, Funding acquisi­
heat deflection temperature (HDT) and the dynamic mechanical analysis tion, Investigation, Methodology, Supervision, Validation, Writing –
(DMA) was presented to enable a faster characterization of 3D printed review & editing.
photopolymerized systems. In comparison to conventional measure­
ment methods of HDT and DMA, the main advantages of combined Declaration of competing interest
technique are faster sample preparation and measurement, lower ma­
terial consumption, time, energy and capacity savings of the instrument The authors declare the following financial interests/personal re­
and the ability to compare the results of DMA and HDT for the same lationships which may be considered as potential competing interests:
sample, while eliminating sample preparation error. Several thermo­ Prusa Polymers company conducts commercial activities in the field of
mechanical properties obtained through this method together with 3D printing including manufacturing and selling of 3D printing equip­
auxiliary techniques were used for evaluating the influence of the pro­ ment, materials and consumables. Among others, it has developed and
cessing and post-processing parameters such as the post-curing time, manufactured the SL1 printer and CW curing unit used within the study.
specimen thickness, or object orientation on the print platform. - M.Z.
It was concluded that print orientation has large effect on both HDT
and temperature at break. The printing in x direction yielded the most Acknowledgement
mechanically resistant arrangement because the applied force was
perpendicular to the printed layers and the weak interlayer regions. The Petr Lepcio acknowledges the GF21-43070L project supported by the
combined DMA-HDT analysis showed that the mechanical properties GA ČR. Martina Štaffová acknowledges the internal student project of
the Central European Institute of Technology, Brno University of

10
M. Štaffová et al. Polymer Testing 108 (2022) 107499

Technology [grant number CEITEC VUT-J-20–6317]. Marek Zbončák [20] Oropallo William, Les A. Piegl, Ten challenges in 3d printing, Eng. Comput. 32 (1)
(2016) 135–148, https://doi.org/10.1007/s00366-015-0407-0. Issn 0177-0667.
acknowledges the FW01010650 project supported by the TA ČR. We
Dostupné z.
also acknowledge the CzechNanoLab Research Infrastructure supported [21] Ali a Bagheri, Jianyong Jin, Photopolymerization in 3d printing, Acs Applied
by MEYS CR (LM2018110) for providing equipment for microscopical Polymer Materials 1 (4) (2019) 593–611, https://doi.org/10.1021/
analysis. acsapm.8b00165. Issn 2637-6105. Dostupné z.
[22] Wiktoria Tomal, Krok Dominika, Chachaj-Brekiesz Anna, Lepcio Petr,
Ortyl Joanna, Harnessing light to create functional, three-dimensional polymeric
Appendix A. Supplementary data materials: multitasking initiation systems as the critical key to success, Addit.
Manuf. 48 (2021), https://doi.org/10.1016/j.addma.2021.102447. Issn
22148604. Dostupné z.
Supplementary data to this article can be found online at https://doi. [23] Hong, Yong Sung, Kim Ye Chan, Mei Wang, et al., Experimental investigation of
org/10.1016/j.polymertesting.2022.107499. mechanical properties of Uv-Curable 3d printing materials, Polymer 145 (2018)
88–94, https://doi.org/10.1016/j.polymer.2018.04.067. Issn 00323861. Dostupné
z.
References [24] Jaroslaw Kotlinski, Mechanical properties of commercial rapid prototyping
materials, Rapid Prototyp. J. 20 (6) (2014) 499–510, https://doi.org/10.1108/Rpj-
[1] Tuan D. Ngo, Kashani Alireza, Imbalzano Gabriele, T.Q. Kate, David Hui Nguyen a, 06-2012-0052. Issn 1355-2546. Dostupné z.
Additive manufacturing (3d printing): a review of materials, methods, applications [25] Kim Ye Chan, Hong Sungyong, Hanna Sun, et al., Uv-curing kinetics and
and challenges, Compos. B Eng. 143 (2018) 172–196, https://doi.org/10.1016/j. performance development of in situ curable 3d printing materials, Eur. Polym. J.
compositesb.2018.02.012. Issn 13598368. Dostupné z. 93 (2017) 140–147, https://doi.org/10.1016/j.eurpolymj.2017.05.041. Issn
[2] Carl G. Schirmeister, Hees Timo, H. Licht a Erik, Rolf Mülhaupt, 3d printing of high 00143057. Dostupné z: doi.
density polyethylene by fused filament fabrication, Addit. Manuf. 28 (2019) [26] Yi Gao, Xu Lei, Zhao Yang, Zhengwei You a Qingbao Guan, 3d printing preview for
152–159, https://doi.org/10.1016/j.addma.2019.05.003. Issn 22148604. stereo-lithography based on photopolymerization kinetic models, Bioactive
Dostupné z. Materials 5 (4) (2020) 798–807, https://doi.org/10.1016/j.
[3] R. Boros, P. Kannan Rajamani a, J.G. Kovacs, Combination of 3d printing and bioactmat.2020.05.006. Issn 2452199x. Dostupné z.
injection molding: overmolding and overprinting, Express Polym. Lett. 13 (10) [27] Uzcategui Asais Camila, Archish Muralidharan, Virginia L. Ferguson, J. Stephanie,
(2019) 889–897, https://doi.org/10.3144/expresspolymlett.2019.77. Issn Bryant a Robert R. Mcleod, Understanding and improving mechanical properties in
1788618x. Dostupné z. 3d printed parts using a dual-cure acrylate-based resin for stereolithography, Adv.
[4] Tianyun Yao, Zichen Deng, Kai Zhang, Li Shiman, A method to predict the ultimate Eng. Mater. 20 (12) (2018), https://doi.org/10.1002/adem.201800876. Issn
tensile strength of 3d printing polylactic acid (Pla) materials with different printing 14381656. Dostupné z.
orientations, Compos. B Eng. 163 (2019) 393–402, https://doi.org/10.1016/j. [28] Mendes Felipe, David Patrocinio Cristian, José M. Laza, Leire Ruiz-Rubio, José
compositesb.2019.01.025. Issn 13598368. Dostupné z. Luis Vilas-Vilela, Evaluation of postcuring process on the thermal and mechanical
[5] Juraj Svatík, Lepcio Petr, Ondreáš František, Zárybnická Klára, Zbončák Marek, properties of the Clear02™ resin used in stereolithography, Polym. Test. 72 (2018)
Josef Jančář Přemysl Menčík a, Pla toughening via bamboo-inspired 3d printed 115–121, https://doi.org/10.1016/j.polymertesting.2018.10.018. Issn 01429418.
structural design, Polym. Test. 104 (2021), https://doi.org/10.1016/j. Dostupné z: doi:.
polymertesting.2021.107405, 01429418. Dostupné z. [29] Ondreas Frantisek, Jancar Josef, Temperature, frequency, and small static stress
[6] Charles W. Hull, Apparatus for Production of Three-Dimensional Objects by dependence of the molecular mobility in deformed amorphous polymers near their
Stereolithography, Us4575330a. Uděleno, 1984. glass transition, Macromolecules 48 (13) (2015) 4702–4716, https://doi.org/
[7] Ligon Samuel Clark, Liska Robert, Stampfl Jürgen, Matthias Gurr a Rolf Mülhaupt, 10.1021/acs.macromol.5b00550. Issn 0024-9297. Dostupné z: doi:.
Polymers for 3d printing and customized additive manufacturing, Chem. Rev. 117 [30] Alexander Bardelcik, Yang Steven, Faraz Alderson, a Andrew Gadsden, The effect
(15) (2017) 10212–10290, https://doi.org/10.1021/acs.chemrev.7b00074. Issn of wash treatment on the mechanical properties and energy absorption potential of
0009-2665. Dostupné z: doi:. a 3d printed polymethyl methacrylate (Pmma), Mater. Today Commun. 26 (2021),
[8] Aydin Sadeqi, Nejad Hojatollah Rezaei, E. Rachel, Owyeung a sameer sonkusale, https://doi.org/10.1016/j.mtcomm.2020.101728. Issn 23524928. Dostupné z.
Three dimensional printing of metamaterial embedded geometrical optics (Mego) 5 [31] Sultan Aati, Akram Zohaib, Barsha Shrestha, Patel Jainish, Shih Benjamin,
(1) (2019), https://doi.org/10.1038/s41378-019-0053-6, 2055-7434. Dostupné z. Shearston Kate, Ngo a Hien, Amr Fawzy, Effect of post-curing light exposure time
[9] Zeyu Chen, Song Xuan, Lei Liwen, et al., 3d printing of piezoelectric element for on the physico–mechanical properties and cytotoxicity of 3d-printed denture base
energy focusing and ultrasonic sensing, Nano Energy 27 (2016) 78–86, https://doi. material, Dent. Mater. (2021), https://doi.org/10.1016/j.dental.2021.10.011. Issn
org/10.1016/j.nanoen.2016.06.048. Issn 22112855. Dostupné z. 01095641. Dostupné z.
[10] Brett E. Kelly, Indrasen Bhattacharya, Heidari Hossein, Shusteff Maxim, [32] I. Gibson, D. Rosen, B. Stucker, Additive Manufacturing Technologies: 3d Printing,
M. Christopher, K. Taylor Spadaccini a Hayden, Volumetric additive Rapid Prototyping, and Direct Digital Manufacturing, second ed., Springer, New
manufacturing via tomographic reconstruction, Science 363 (6431) (2019) York, 2015, ISBN 978-1-4939-2112-6.
1075–1079, https://doi.org/10.1126/science.aau7114. Issn 0036-8075. Dostupné [33] Benjamin D. Fairbanks, F. Timothy, Christopher J. Kloxin Scott, S. Kristi, N. Anseth
z. a Christopher, Thiol− Yne Bowman, Photopolymerizations: novel mechanism,
[11] Xiaoyan Xu, Awad Atheer, Pamela Robles-Martinez, Gaisford Simon, W. Alvaro kinetics, and step-growth formation of highly cross-linked networks,
Goyanes a Abdul, Basit. Vat photopolymerization 3d printing for advanced drug Macromolecules 42 (1) (2009) 211–217, https://doi.org/10.1021/ma801903w.
delivery and medical device applications, J. Contr. Release 329 (2021) 743–757, Issn 0024-9297. Dostupné z.
https://doi.org/10.1016/j.jconrel.2020.10.008. Issn 01683659. Dostupné z: doi:. [34] Melick Van, H.G.H.L.E. Govaert, H.E.H. Meijer, On the origin of strain hardening in
[12] Sophia N. Economidou, Dimitrios A. Lamprou a Dennis Douroumis, 3d printing glassy polymers, Polymer 44 (8) (2003) 2493–2502, https://doi.org/10.1016/
applications for transdermal drug delivery, Int. J. Pharm. 544 (2) (2018) 415–424, S0032-3861(03)00112-5. Issn 00323861. Dostupné z.
https://doi.org/10.1016/j.ijpharm.2018.01.031. Issn 03785173. Dostupné z: doi:. [35] Xiangquan Wu, Lian Qin, Li a Dichen, Zhongmin Jin, Tilting separation analysis of
[13] Bloomquist, J. Cameron, Michael B. Mecham, Mark D. Paradzinsky, et al., bottom-up mask projection stereolithography based on cohesive zone model,
Controlling release from 3d printed medical devices using Clip and drug-loaded J. Mater. Process. Technol. 243 (2017) 184–196, https://doi.org/10.1016/j.
liquid resins, J. Contr. Release 278 (2018) 9–23, https://doi.org/10.1016/j. jmatprotec.2016.12.016. Issn 09240136. Dostupné z: doi:.
jconrel.2018.03.026. Issn 01683659. Dostupné z. [36] J. Jancar, W. Wang, A.T. Dibenedetto, On the heterogeneous structure of thermally
[14] Pant, Prabhat. Residual stress distributions in additively manufactured parts: effect cured bis-Gma/Tegdma resins, J. Mater. Sci. Mater. Med. 11 (11) (2000) 675–682,
of build orientation. Dostupné z: doi:10.3384/lic.diva-164108. https://doi.org/10.1023/A:1008999023271. Issn 09574530. Dostupné z:.
[15] Yuxuan Wang, Yonghui Zhou, Lin Lanying, Jorge Corker a Mizi Fan, Overview of [37] D.K. Chattopadhyay, Siva Sankar Panda, K.V.S.N. Raju, Thermal and mechanical
3d additive manufacturing (Am) and corresponding Am composites, Compos. Appl. properties of epoxy acrylate/methacrylates Uv cured coatings, Prog. Org. Coating
Sci. Manuf. 139 (2020), https://doi.org/10.1016/j.compositesa.2020.106114. Issn 54 (1) (2005) 10–19, https://doi.org/10.1016/j.porgcoat.2004.12.007. Issn
1359835x. Dostupné z. 03009440. Dostupné z: doi:.
[16] C. Hofstetter, S. Orman, S. Baudis, J. Stampfl, Combining cure depth and cure [38] Cosmi Francesca, Dal Maso Alberto, A mechanical characterization of Sla 3d-
degree, a new way to fully characterize novel photopolymers, Addit. Manuf. 24 printed specimens for low-budget applications, Mater. Today Proc. 32 (2020)
(2018) 166–172, https://doi.org/10.1016/j.addma.2018.09.025. Issn 22148604. 194–201, https://doi.org/10.1016/j.matpr.2020.04.602. Issn 22147853. Dostupné
Dostupné z. z: doi:.
[17] Chunguang Li, Jue Cheng, Wenkai Chang, Nie Jun, Photopolymerization kinetics [39] Joseph Borrello, Nasser Philip, C. James, Iatridis a Kevin D. Costa, 3d printing a
and properties of a trifunctional epoxy acrylate, Des. Monomers Polym. 16 (3) mechanically-tunable acrylate resin on a commercial Dlp-Sla printer, Addit. Manuf.
(2012) 274–282, https://doi.org/10.1080/15685551.2012.747143. Issn 1568- 23 (2018) 374–380, https://doi.org/10.1016/j.addma.2018.08.019. Issn
5551. Dostupné z. 22148604. Dostupné z.
[18] Young-Jun Park, Lim Dong-Hyuk, Kim Hyun-Joong, Park a Dae-Soon, Sung Ick- [40] Dohyun Kim, Shim Ji-Suk, Lee Dasun, Shin Seung-Ho, Nam Na-Eun, Park Kyu-
Kyung, Uv- and thermal-curing behaviors of dual-curable adhesives based on epoxy Hyung, June-Sung Shim, Kim Jong-Eun, Effects of post-curing time on the
acrylate oligomers, Int. J. Adhesion Adhes. 29 (7) (2009) 710–717, https://doi. mechanical and color properties of three-dimensional printed crown and bridge
org/10.1016/j.ijadhadh.2009.02.001. Issn 01437496. Dostupné z: doi:. materials, Polymers 12 (11) (2020), https://doi.org/10.3390/polym12112762.
[19] Stansbury Jeffrey, W. a Mike, J. Idacavage, 3d printing with polymers: challenges Issn 2073-4360. Dostupné z.
among expanding options and opportunities, Dent. Mater. 32 (1) (2016) 54–64,
https://doi.org/10.1016/j.dental.2015.09.018. Issn 01095641. Dostupné z:.

11