3.1. Characterization on the Neat Starting Materials and the Composite One
A full characterization of the neat starting materials, compared to the Na–CMC/PDMS composite, has been firstly performed and results are shown in Figure 1
. Specifically, FTIR and XRD analysis have been carried out in order to investigate the chemical composition and structure of the employed materials, TGA and DSC thermograms were acquired to characterize their thermal properties, while rheological characterization of PDMS pellets has been performed as a preliminary step to the polymer-extrusion and 3D-printing processes.
In Figure 1
a it is possible to notice the FTIR spectra of the different materials, while in Figure S1 of the Supporting Information file
, the spectra of PDMS and cellulose powder, with their characteristic peaks, are shown in detail. The typical peaks of –CH3
symmetric rocking and Si–C symmetric stretching in (Si–CH3
) lie in the range 789–796 cm−1
. The asymmetrical (Si–O–Si) stretching vibrations are visible at about 1020–1074 cm−1
, while the –CH3
symmetric bending in (Si–CH3
) is evident at about 1260 cm−1
]. The peak at about 1411 cm−1
is ascribable to the –CH3
asymmetric bending in (Si–CH3
), as confirmed in the literature [19
]. The stretching vibrations of C=C and CH–(CH2
) are also visible at 1632 cm−1
and 2866 cm−1
, respectively. Then, the peaks at about 2801 and 2905 cm−1
are representative of the symmetrical and asymmetrical –CH3
stretching, respectively [19
]. The presence of a broad peak at 3500 cm−1
, typical of –OH functional group, is most probably due to adsorbed moisture (Figure S1a in Supporting Information file
The characteristic peaks of the employed Na–CMC powder are: at about 1100 cm−1
the one typical of n(C–O) is identifiable, while at 1622 cm−1
the carboxyl groups n(COOH) stretching vibration are represented. Then, the peaks located at 2929 cm−1
and 3400 cm−1
correspond to n(CH2
) and n(OH) stretching vibrations [20
], respectively. The peak at about 3500 cm−1
is representative of –OH groups in the Na–CMC chemical structure and those due to the adsorbed moisture (Figure S1b in Supporting Information file
). From Figure 1
a it is evident that the spectrum of the composite and that of the pure PDMS are nearly the same. The diffraction patterns, shown in Figure 1
b, highlight the presence of two different peaks for pure PDMS as confirmed in the literature [22
]: the first one, exhibiting bigger amplitude, is located at around 11.65°, while the second one, smaller and broader, lies at 20.68°, thus suggesting an amorphous microstructure of the polymer [23
]. Na–CMC exhibits a broad peak centered at 20°, representative of the low crystallinity degree of the Na–CMC structure. Therefore, the starting materials employed in this study are all defined as amorphous. In the structure of the composite material, the two peaks, at 11.65° and 20°, are both observed. The results of the thermal analysis are shown in Figure 1
c,d, where TGA and DSC thermograms are respectively shown.
The TGA curve related to PDMS exhibits a main step, occurring between 330 °C and 630 °C and representative of the degradation of the polymer-silicone backbone. An initial, slight weight loss, corresponding to approximately 0.3% of the initial mass, is due to moisture content. After 630 °C, the residual mass is minimal, around 0.01%, indicating that decomposition products are volatile.
TGA signal of Na–CMC, shown in detail in Figure S2 (in the Supporting Information file)
, reveals four different degradation steps, explained by the fact that Na–CMC is a complex copolymer composed by different chemical groups. The first step, completed at about 250 °C and corresponding approximately to 12% of weight loss, is due to moisture evaporation; the second weight loss, the most considerable (around 62.5%), lies in the range 200–550 °C (inflection point at 289.13 °C) and is representative of the degradation of the saccharide rings, the breaking of C–O–C bonds in the CMC chain, and the elimination of CO2
from the polymeric backbone [24
]; the third weight loss, around 21.5%, exhibits an inflection point at 697.66 °C and corresponds to the degradation of further organic material; while the last step, with a weight loss of about 11% and an inflection point at about 935.36 °C, is representative of sodium evaporation. In this case, a minimal residual mass is also proved in the end of the measurement, indicating that the decomposition products are volatile.
From Figure 1
c it is evident that the presence of cellulose in the composite filament anticipates the beginning of the softening and degradation processes. It also modifies the rate of degradation of PDMS matrix, as denoted by the higher slope of the Na–CMC/PDMS curve when compared to the bare PDMS one. On the contrary, from Figure 1
d it emerges that the cellulose presence in the composite does not cause significant modifications in the DSC thermogram. DSC thermograms of the pure starting materials, shown in Figure S3 (Supporting Information file)
, reveal that PDMS softens rather than melts, confirming its amorphous character, while Na–CMC appears to be a highly hygroscopic material.
In order to fix the working temperatures of the extruder chambers, used to shape PDMS pellets into filaments successively processed by the 3D printer, a preliminary rheological characterization has been carried out on PDMS pellets, as reported in Figure 2
A dynamic rheological ramp test has been firstly performed in Figure 2
a, where the trend of viscosity with respect to temperature variations and of G modules with respect to time have been evaluated, in order to assess the minimum processing temperature and a possible degradation of the material could suffer during the extrusion process. The test has been performed by varying temperature from 100 °C to 205 °C, according to the recommendation for extrusion in the polymer datasheet. The obtained results, shown in the graph, highlight a viscoelastic behavior of PDMS pellets. The intersection of storage and loss modules, G′ and G″, corresponding to the complete material softening, occurs at a temperature of about 170 °C, a value in agreement with the “melting range” reported in the material datasheet (170–205 °C). The viscosity increase, recorded in the beginning of the test, is most likely due to the thermal expansion to which the material is subjected and is a typical behavior of viscoelastic materials, like polymers. After its maximum, reached at 127 °C, viscosity starts to decrease as the material softens, and continues until the end of the test, when temperature reaches the value of 205 °C, the upper limit of the indicated “melting range”. After repeated rheological tests (data not shown) were performed to fix the working temperature of the different extruder chambers, two temperatures have been selected, 190 °C and 205 °C, at which rheological isothermal analysis has been then performed (Figure 2
b,c). From the comparison of Figure 2
b,c it emerges that by increasing the temperature, polymer softening occurs before, respectively after 34.49 min at 190 °C, and after 12 min when kept at 205 °C.
d is representative of the viscosity as a function of the shear rate, at three specific temperature values: the obtained softening temperature of 170 °C, and those chosen as the extruder working temperatures, that are 190 °C and 205 °C. All the acquired curves exhibit the characteristic behavior of pseudoplastic materials, typical of many polymers in the melt/soften state, and, while increasing temperature, the viscosity values decrease, as expected.
3.3. Rheological Properties of PDMS and Na–CMC/PDMS Filaments
A comparison between the two kinds of filaments rheological properties has been then carried out in order to investigate the influence of the cellulose addition on the PDMS processability in the further FDM step and assess the working parameters of the 3D printer.
First of all, isothermal analysis has been performed at 190 °C (Figure 4
a,b) and 205 °C (Figure 4
c,d), the same temperatures at which PDMS pellets were previously characterized and then extruded, as the already extruded filaments have to undergo to a further “ideal” extrusion process in the printer. Moreover, the same characterization has been performed at 230 °C (Figure 4
e,f) that is the temperature that will be finally chosen for the 3D-printing step, as deeply explained in the Section 3.4
“Filament 3D printing”. The results are reported in Figure 4
, where the graphs in the first column are referred to the neat PDMS filament, while those in the second one to the Na–CMC-modified PDMS filament. The graphs show G′ and G″ modules as a function of time and the viscosity curve recorded at the specified temperature. The crossover point, at the intersection between G′ and G″, and the corresponding viscosity values have been highlighted in each graph. In both filament cases, when temperature is increased, the related timespan time is reduced, as observed also in the case of pure pellets. Moreover, the addition of cellulose fibrils in the PDMS matrix moves up the occurrence of the material softening, as the inflection point lies at shorter time interval in graphs of Figure 4
b,d, rather than in that of Figure 4
The main data of the described rheological analysis are summarized in Table 2
From data summarized in Table 2
it can be deduced that, for a fixed temperature value, melting time decreases (the intersection of G′ and G″ modules occurs before) moving from PDMS pellets to PDMS filament. This is most probably due to the fact that PDMS filaments have been already thermally treated during the previous extrusion process in order to acquire their final shape. The presence of cellulose facilitates even more the silicone-melting process; indeed, in the case of the Na–CMC/PDMS filament, the intersection of modules occurs before those of PDMS pellets and filament. At the same time, moving from PDMS pellets to pure and composite PDMS filament, the modulus and viscosity values decrease. Therefore, the presence of cellulose causes a reduction in viscosity. Moreover, from Table 2
it emerges that, for each type of sample, a temperature increment causes the crossover point to move up and a reduction of the modulus and viscosity values.
Further rheological tests have been carried out on the filaments in order to characterize their behavior during the printing process, at the temperatures of interest, and results are shown in Figure 5
is representative of the variation of viscosity when varying the shear rate for both PDMS and Na–CMC/PDMS filaments, in correspondence with the softening temperature found for PDMS pellets, the two selected for the extrusion process and the one that will result as the most suitable for the printing (170, 190, 205, and 230 °C, respectively). All the examined curves show the typical trend of a pseudoplastic material, typical of many polymers in the melt state.
The shear rate value at which filaments are subjected to during the 3D-printing process is calculated: for a typical 3D-printing speed of 2 mm/s and an obtained filament diameter of 1.6 mm, dependent on the nozzle size, a shear rate of 1.25 s−1
is calculated as their ratio. At this value, moving from lower to higher temperatures the viscosity decreases, and the same trend is evident also in the case of bare PDMS pellet viscosity (Figure 2
d,e). Moreover, the composite filaments always exhibit lower viscosity values, as also shown in Figure 2
e. Interestingly, in the case of the composite filament, the viscosity values at different shear rates and temperatures (in the range of 3D-printer working conditions) show the same trend of PLA-composite filaments, traditionally employed in FDM applications [26
3.4. Filament 3D Printing
Once the composite filament has been produced by extrusion and deeply characterized, also in comparison with the pure PDMS one, it is tested in FDM. Despite the numerous attempts of 3D printing at temperature values of 190 °C and 205 °C, and related time interval before material degradation, no satisfying results have been obtained. Indeed, the evident surface roughness, larger diameter, and the faster degradation of the composite filament caused nozzle obstruction, which prevented the printing process completely, or led to the printing of a degraded filament. Therefore, further combinations of temperature and related span time have been tested, and, among these, the one obtained setting a temperature of 230 °C, with a corresponding span time of 2.74 min (3.92 for the pure PDMS filament), even if too short, resulted to be effective in 3D printing; indeed, lower temperature values did not permit to print at all. In this way, a first attempt of FDM of the composite PDMS filament has been performed, as a primary step for the production of three-dimensional objects, such as a model of human heart. Figure 6
shows photos of the satisfying filament used for FDM (Figure 6
a) and micrographs of their deposition in layers (Figure 6
b–d), where it is possible to notice that layers adhere between each other completely, and the interfaces are pointed at by the arrows. Nevertheless, highly irregular samples have been obtained (see Figure S4
), as the employed 3D printer did not permit a fine control of geometry while using the current material. This is most probably due to the intrinsic nature of the used polymeric matrix, which does not properly melt, but softens and degrades very quickly. On the other hand, PDMS-like polymers are often successfully 3D-printed by making changes in the traditional FDM experimental setup, for example, by switching the basic filament-dispensing system of an FDM machine with a syringe and needle, allowing them to avoid modifications in the stage, software, or the printing unit [27
]. Hinton et al. [28
] used a hydrophilic support bath via freeform reversible embedding to extrude PDMS within the hydrophilic Carbopol gel that is later cured by heating in two rounds, managing to create perfusable manifolds using this technique. Li and group printed wax instead of PDMS by replacing the syringe with a glass nozzle [29
]. Subsequently, they use these wax prints as molds for casting PDMS. This approach results to be better than direct-printing PDMS, as it obviates the need of employing a complicated gel matrix-based printing, leading to smoother surfaces at the same time as wax reduces the process-introduced roughness.
Hence, when using a silicon-based polymer, also modified with a cellulose filler, the use of different and more robust 3D printers, and/or a modified printing process are strongly recommended.