3.1. General Issues
Figure 3 shows the cylindrical substrates manufactured with and without Cu NWs at different concentrations (see
Table 1), where the flat parallel surfaces (first and last manufactured layer) and their respective contours are well defined. However, it can be observed that the lateral surface of the substrates manufactured with Cu NWs, especially those of the one with a
, present some irregularities or deformations, possibly because the laser with a
wavelength does not photopolymerize in perfect conditions, due to the presence of the additives. In principle, due to the high proportion of IPA that the Cu NWs solution contains, the curing time of each polymerized layer, whose superposition leads to the final object, should increase to compensate for the energy scattered due to the additives. Additionally, it is important to point out that the post-curing time, carried out at room temperature, should be longer when the IPA concentration increases, so as to compensate the lack of polymerization if the stereolithography machine’s working parameters cannot be modified, as is our case.
Table 2 reports the values of key measurements (length, mass and volume) of the cylindrical substrates, obtained from the CAD design and checked directly after manufacturing. The values show slight mismatches between the measurements obtained from the CAD file and from the actual prototypes (slightly higher). Furthermore, both the volumetric mismatch and mass decrease as the concentration of Cu NWs increases. It is likely that a significant part of the IPA contained in the resin evaporates during the
-min mixing due to its high vapor pressure (
at
). The rest may follow an esterification reaction with the methacrylic acids contained in the resin [
18]. As the load of Cu NW solution increases in the resin, the samples prepared by SLA reach a lower final mass and volume. So, on the one hand, a higher content of Cu NW solution could lead to increased evaporation of IPA and to a higher decrease in mass (
at
and
at
). On the other hand, a higher content of Cu NW solution would leave more residual IPA in the resin, leading to a higher rate of esterification, and hence to a more significant shrinkage of the polymer volume (
at
and
at
) [
19]. Finally, as the decrease in mass is more important than the decrease in volume, the density increases by
and
, when the load of Cu NW solution increases to
and
, respectively. This can be taken into account for design purposes if precise geometrical requirements upon final parts are needed.
The manufacturing time of a specimen, cylindrical or for tensile testing, with or without Cu NWs (see fabrication parameters in
Table 1), is approximately
and
seconds respectively. Then, for the cylindrical substrate on average each layer, with a surface area of
, is manufactured in approximately
s. The cost of manufacturing a functional device with Cu NWs, depending on the percentage of Cu NWs added, increased approximately among a
and a
, when compared to the non-functional substrate. For instance, the cost of each specimen without Cu NWs is approximately
, while that of those manufactured at
and
are
and
respectively. It is very important to keep in mind that this cost only considers the chemicals used for synthesis and washing of the NWs, and not the energy, the inert gas used and other complementary supplies for manufacturing. In
Table 3 the approximate cost of materials used for each manufactured object or test probe is reported.
3.2. Morphology
The surface morphology and presence of elements in the fabricated substrates with Cu NWs at
(
Figure 3b and
Figure 4a) and
(
Figure 3c and
Figure 4b), are quantitatively and qualitatively analyzed by scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS). The micrographs in
Figure 4a,b show well-dispersed Cu NWs in the photopolymerized resin. It is clear from these observations that the Cu NWs content is well below the percolation threshold. Consequently, the Cu NWs nanofillers are expected to influence the thermo-mechanical properties [
20,
21] of the polymer matrix, but not its electrical properties [
22]. The spectra numbered with
to
, in
Figure 4a,b, are the regions where the Cu NWs were found. The spectra numbered with
are the regions without Cu NWs in both
Figure 4a,b.
Table 4 and
Table 5 reports the identified elements, in atomic percent (
), measured in the regions with and without Cu NWs. Additionally, according to the laser direction that polymerizes the resin during SLA printing, the surface roughness can vary significantly by adding the nano-additives and a homogenous filler distribution should be pursued to achieve an adequate dispersion for adequate manufacturability and functionalization (i.e., change of the mechanical and electromagnetic properties). In spite of possible improvements, the manufacturing of three-dimensional objects by laser stereolithography, using a Cu NWs mass-functionalized photopolymer, is demonstrated.
In
Figure 4c,d, the X-ray computed tomography (CT) is used to capture the distribution and identification of the Cu NWs in the bulk of substrates manufactured in photoreactive resin through SLA technique by 3D printing. The brightness of X-ray CT images depends on the amount of X-ray penetration, which allows the identification of several Cu NWs randomly distributed into the bulk because, as the density of the specimens increases, the amount of X-ray penetration decreases, resulting in a brighter image [
23,
24,
25]. In the substrate section analyzed, the bulk distribution of Cu nanowires is identified and corresponds approximately to a
of the
volume analyzed.
3.3. Mechanical and Thermal Performance
Representative load-displacement curves for the substrates manufactured using the photoreactive commercial resin, both with and without Cu NWs used for mass-functionalization, are provided in
Figure 5. It can be observed that the curves without Cu NWs and with Cu NWs at
have ductile behaviors, while the
functionalization leads to a brittle behavior.
Figure 5 and
Figure 6 show that the tensile strength significantly increases in the specimen with Cu NWs at
. Mechanical performance decreases in the substrate with Cu NWs at
.
Tensile strength values are
,
and
for the specimen with Cu NWs contents of
,
and resin without nanofillers, respectively. For these samples, corresponding Young’s moduli are
,
and
as shown in
Figure 6, respectively. The increase in mechanical performance is attributed to the successful load transfer from the matrix to the Cu NWs. The decrease in tensile strength above a Cu NWs content of
may be related to the decrease in volume and mass when the content of Cu NWs increases, due to the IPA evaporation during cured and post-cured process. It could also be attributed to a decrease in the interfacial interactions between the polymeric matrix and fillers, due to solubility parameter that determined the substances affinity among the dissolvent and polymer, which do not differ in more than one or two units [
26,
27].
In addition, the presence of fillers may also affect the polymerization degree of the final device producing a plastification effect for the higher filler contents, which should always be taken into account in photopolymerization-based additive manufacturing techniques. Besides, during the strain of specimens, we determined that the molecular structure of the substrates without Cu NWs was the one capable of absorbing more energy, while the probe containing a Cu NWs at proved the worst in terms of energy absorbance. Furthermore, the specimens with Cu NWs at and showed more brittle fracture, when compared to the reference substrate.
Figure 7 shows the surface distribution of the values of storage or reduced modulus (
) and hardness (
), obtained by instrumented nano-indentation as explained in the materials and methods section.
Figure 7 highlights that the variations in
and
, when comparing the measurement performed on the substrates with Cu NWs and without Cu NWs, are not completely uniform.
Figure 7 exhibits some areas with less and greater than the average deformation resistance values (see also
Figure 8), in case of
the minimal value acquired for Cu NWs at
(
Figure 7b) is major to the average value obtained in the pure resin (
Figure 7a), while that the
values minimum, maximum and averages measured in the substrate surface with Cu NWs at
are lower than obtained in the substrate without Cu NWs (
Figure 8). In
Figure 7, a strong correlation between
and
is shown, where the areas probably reflect localized concentrations of Cu NWs than can be comparable in size to the hundreds of micrometer or nanometric-scales. Removal of such entangled agglomerates is a major focus of the many methods [
28,
29,
30,
31,
32] used to disperse nanofillers in polymeric matrices. In general, when the agglomerates are only weakly infiltrated by the polymer, final properties of the composite are not as remarkable, as might be achieved by well-dispersed nano-additives bound to the matrix. This degradation of properties probably corresponds to the local “soft” spots, while the local “hard” spots probably reflect enhanced areas of Cu NWs concentration with adequate polymer infiltration (
Figure 7). The advantage of such maps (
Figure 7) is that local variations in properties are assessed directly and do not have to be inferred from measurements of entire composite components [
31,
32].
Figure 8 presents the average values of E’ and H for the pure resin and for the one loaded with the
Cu NW solutions. The increase in storage modulus when loading the resin with Cu NWs can be once again attributed to the load transfer while the decrease in hardness could be due to the decrease of the density related to the evaporation of IPA during the UV curing of the resin.
Accordingly, if globally averaged properties are required for a polymeric or composite matrix to predict the overall response of a device, then the response of a number of indentations over a large enough area is required. Taking into account practical considerations, the indentation-based mapping of the mechanical properties of polymeric systems (soft materials) will always require indentation spacing greater than that employed for harder materials. Therefore, matching the indentation spacing and size to the length scale of the microstructure is important.
Figure 9 shows the change in
and
for the substrates manufactured using the photo-resin without Cu NWs, with Cu NWs at
, and
, as a function of temperature.
Figure 9a shows that
of the substrate without Cu NWs has the highest value,
, and decreases with the addition of nanowires. The lowest value,
, was obtained for the substrate with Cu NWs at
. Accordingly, the factors such as the reaction degree and cross-linking density mainly influence the value of
. Besides this, the curves clearly highlight the glass transition region. Therefore, a distributed glass transition process in the material is confirmed, suggesting that the material behavior can be evaluated through rheological properties.
Figure 9b shows the change of
of the substrates with Cu NWs with respect to those without Cu NWs. The peak value of
is mainly influenced by the glass transition temperature (
). It can be appreciated that the increase of Cu nanowires, and the consequent lower cross-linking density, leads to higher mobility of polymer chains during glass transition and to a lower
value. The sharper peak of
indicates that a more regular structure is formed with the increasing content of Cu NWs. In these tests, the
in the substrates without Cu NWs and with the loading of
and
reaches
,
, and
respectively.
The results of the DSC tests reported in
Figure 10a, show for the first scan a broadly distributed second glass transition region and the presence of an exothermal signal. The distribution of the second glass transition region seems to cover regions: between
for the substrate without Cu NWs, between
for the substrate with
Cu NWs and between
for the substrate with
Cu NWs. Thus, the
value slightly decreased as the Cu NWs loading in resin increased.
The thermal stability of the samples was investigated by TGA to ensure that the nano-additives are stable. Under pyrolytic conditions in the
atmosphere, the degradation of the samples occurred with a sharp weight loss around
to
, accompanied with possible evolved organic fragments (e.g., methacrylic acid, ester, etc.) [
33,
34]. TGA results between
and
are provided in
Figure 10b. Slight mass loss from TGA curves, around
was attributed to the loss of absorbed IPA in resin, where most of the IPA to be evaporated during the UV post-curing, and moreover in general, between
and
, you lose the physisorbed water. Effects of Cu NWs on the thermal degradation behavior of the nanocomposites can be easily seen within the DTGA curves provided in
Figure 10c. The thermal degradation temperatures (
), along with fusion temperature values (
) and glass transition temperature (
) of the substrates with and without nano-additives are tabulated and provided in
Table 6. The fusion and degradation temperature are almost unchanged with the addition of the Cu NWs. One of the reasons for this behavior is the lack of chemical interaction between the resin and the Cu NWs. In fact, the degradation leads to a general breakage of C–C bonds, reducing the chemical cross-linking points and thus to an increase of the mobility of the polymeric chains which corresponds to a decrease of glass transition temperature. Accordingly, before the resin starts to degrade, surface interactions might get lost between the resin and the Cu NWs [
34,
35].
3.4. Spectroscopic Characterization
The Raman spectra obtained on the substrates manufactured in the “clear FLGPCL 02” resin without, with
, and with
Cu NWs are shown in
Figure 11. The spectrum of
Figure 11a reveals three Raman burly peaks at
(strong),
and
(very strong) that corresponded to the
bond vibrations, for the surface without Cu NWs. The peak (weak) at
is attributed to the
bond vibration, and the peaks (medium) at
and
are attributed to the
bond vibration. The band between 800 and
correspond to the
bond vibrations, between
and
to the
bond vibrations, and between
and
to the
bond asymmetric stretching vibrations. In
found the
bond asymmetric stretching vibration and in
the
bond vibration. Additionally, the weak peaks at
, and
are attributed to the
and
bonds vibrations respectively.
Raman spectroscopy reveals the surface interaction between polymeric organic materials and metallic Cu NWs. Consequently, the use of metal dopants such as the Cu NWs in the structure of organic materials allows, through surface-enhanced Raman scattering (SERS), may allow us to acquire the intrinsic vibrational fingerprint of photoreactive commercial resins more intense and defined bands, due to the extremely high sensitivity provided by plasmonic nanomaterials as such as the Cu, as is shown in
Figure 11b,c [
36,
37,
38]. In accordance with the above, this surface interaction can be responsible for the difference of
before and after doping. Additionally, these studies allow us to identify differences in the degree of crystallinity due to the influence of the doping agents.
Figure 12 shows the FT-IR spectrum of the substrates manufactured using the photo-resin with and without nano-additives, where distinct absorption bands from
to
appear, which can be attributed to the
stretching vibration. The two bands at
and
can be attributed to the
-methyl group vibrations. Moreover, the bands at
,
,
and
can be attributed to the
,
,
and
stretching vibration, respectively. The band at
shows the presence of the ester group (
stretching vibration). The band at
can be attributed to the bending vibration of the
bonds of the
group. The two bands at
,
and
can be assigned to the
bond stretching vibrations. Furthermore, there are two weak absorption bands at
and
, which can be attributed to the
group stretching and bending vibrations, respectively. Notwithstanding this, there are no appreciable differences between the FTIR spectra of the resin with and without Cu NWs, because the resin does not form any covalent bond with this type of nano-additives. Additionally, the Cu NWs do not add new functional groups that can be observed with infrared techniques (as appreciated from
Figure 12b,c). Summarizing the above discussions, it can be concluded that the photoreactive polymer resin is macromolecular of type methacrylic acid (MAA) and methyl methacrylate (MMA) [
35,
39,
40].