Influence of Crystallization Kinetics and Flow Behavior on Structural Inhomogeneities in 3D-Printed Parts Made from Semi-Crystalline Polymers

We report the results of a study focusing on the influence of crystallization kinetics and flow behavior on structural inhomogeneities in 3D-printed parts made from polyamide 12 (PA12) and poly(lactic acid) (PLA) by dynamic mechanical analysis (DMA), differential scanning calorimetry (DSC), fast scanning calorimetry (FSC), and wide-angle X-ray diffraction (WAXD). Temperature-dependent WAXD measurements on the neat PLA filament reveal that PLA forms a single orthorhombic α phase during slow cooling and subsequent 2nd heating. The PA12 filament shows a well pronounced polymorphism with a reversible solid–solid phase transition between the (pseudo)hexagonal γ phase near room temperature and the monoclinic α′ phase above the Brill transition temperature TB = 140 °C. The influence of the print bed temperature Tb on structure formation, polymorphic state, and degree of crystallinity χc of the 3D-printed parts is investigated by height and depth-dependent WAXD scans and compared with that of 3D-printed single layers, used as a reference. It is found that the heat transferred from successive layers has a strong influence on the polymorphic state of PA12 since a superimposed mixture of γ and α phases is present in the 3D-printed parts. In the case of PLA, a single α phase is formed. The print bed temperature has, in comparison to PA12, a major influence on the degree of crystallinity χc and thus the homogeneity of the 3D-printed parts, especially close to the print bed. By comparing the obtained results from WAXD, DMA, DSC, and FSC measurements with relevant printing times, guidelines for 3D-printed parts with a homogeneous structure are derived.

Temperature-dependent structure formation in the bulk state for PA12 and PLA laments PA12. Figure SI1 presents the azimuthal integrated 1D scattering pattern of PA12 obtained during step wise cooling and 2 nd heating run.Crystallization starts from 160 °C.At 150 °C, two Bragg reections indexed as q 200 (q 200 = 1.44 Å −1 ) and q 010 (q 010 = 1.48 Å −1 ), corresponding to the α ′ phase 17 (monoclinic unit cell), are observed in the WAXD region during cooling.The corresponding d-spacing are d 200 = 4.36 Å and d 010 = 4.25 Å, respectively.
Upon further cooling, q 200 is shifting to higher q-values, resulting in a merger of q 200 and q 010 at 130 °C to a single sharp reection at q 100 = 1.51 Å −1 (d 100 = 4.15 Å).The single strong reection at ≈ 1.5 Å −1 is associated to the (pseudo)hexagonal γ phase, 47 observed between 130 °C ≤ T ≤ 30 °C.In the low q region two reections are observed at q 002 = 0.42 Å −1 and q 004 = 0.80 Å −1 .The (002) along with it's higher order (004) planes correspond to the periodicity of the amide bonds along the main chain.The scattering pattern of the 2 nd heating run of PA12 are qualitatively similar.In order to obtain the temperature-dependent locations q hkl as well as area A hkl of the Bragg reections and amorphous halo A amo a detailed peak analysis is applied.Representative examples of the deconvoluted 1D scattering pattern in the WAXD region are shown in Figure SI2 for dierent polymorphic states.In Figure SI3 1.0 1.5 2.0 Intensity / a.u.28.95 Å) of PLA. 13 In order to obtain the locations q hkl as well as area A hkl of the Bragg

Results from Dynamic Mechanical Analysis
Terminal relaxation times τ R as well as horizontal shift factors a T from a master curve construction are compiled in Table SI1 for the PA12 and PLA laments.The additional features seen in the POM images either as vertical or horizontal grooves are cutting artifacts resulting from microtomy.The variation of the direction of the surface artifacts in Figure SI10(c,f) is due to changes in the clamping direction.The features are not related to internal structures formed during 3D-printing.Depth dependent POM images captured at depths x of 1.5 mm and 5.0 mm support the ndings reported for the thin lms taken from a depth x of 2.5 mm.In gure SI11 POM images measured on thin lms taken in a depth x of 1.5 mm and 5.0 mm from components printed at dierent bed temperatures

Figure SI 1 :
Figure SI 1: Temperature-dependent azimuthal integrated 1D scattering pattern of PA12 during (a) step wise cooling from the relaxed melt and (b) 2 nd heating run.The temperature interval was between 30 °C and 220 °C, the rate used was ±10 K/min between each measurement step.Major reections are indicated.

Figure SI 3 :
Figure SI 3: Temperature-dependence of the (a) d-spacing and (b) degree of crystallinity χ c during cooling (open symbols) and subsequent 2 nd heating run (full symbols).The degree of crystallinity χ c of as received PA12 at room temperature is shown as red square.The Brill transition temperature T B is indexed with a dashed line.

Figure SI 4 :
Figure SI 4: Temperature-dependent azimuthal integrated 1D scattering pattern of PLA during (a) step wise cooling and (b) subsequent 2 nd heating run.The temperature interval was between 30 °C to 180 °C, the rate between the measurement steps was ±10 K/min.The two major reections are indexed.Further weak reections are indexed in Figure SI5.

Figure SI 5 :Figure SI 6 :
Figure SI 5: Representative example of peak deconvolution for the 1D scattering pattern of PLA in the orthorhombic α phase after slow cooling from the molten state.The baseline corrected 1D scattering data is shown in black, Bragg reections in green, amorphous halo in blue and cumulative t in red dashed.All reections are indexed

Figure
FigureSI7presents the deconvoluted scattering pattern of FFF printed PA12 and PLA components.While the scattering pattern of PLA are qualitatively similar to that obtained during step wise cooling, clear dierences can be observed in case of PA12.Besides the strong reection at ≈ 1.5 Å −1 , corresponding to the (pseudo)hexagonal γ phase, two shoulders at lower and higher q-values occur.A peak deconvolution reveals two additional Bragg reections, most likely caused by a mixture between γ and α phases.

Figure SI 7 :Figure
Figure SI 7: Representative examples of deconvoluted scattering pattern of 3D printed components made of (a) PA12 and (b) PLA.The baseline corrected 1D scattering data is shown in black, Bragg reections in green, amorphous halo in blue and cumulative t in red dashed.

Figure
FigureSI9shows images with 10 k magnication of thin lms microtomed at dierent depth from PA12 components printed at bed temperatures T b of 30 °C, 80 °C and 120 °C.For a depth x of 1.5 mm one can clearly see an improvement of homogeneity with increasing bed temperature T b , i.e. the interfaces between layers become less pronounced.In addition the homogeneity is improving by approaching the core of the component.At a bed temperature T b of 30 °C, the interfaces between individual layers are still visible at a depth x of 5.0 mm.A direct comparison of the images for the components printed at bed temperatures T b of 80 °C or 120 °C demonstrates a higher degree of homogeneity, where the interfaces between printed layers are no longer visible.Components printed at a bed temperature T b of 120 °C

Figure
Figure SI 9: POM images (magnication 10k) of PA12 lms taken from components printed at bed temperatures T b of 30 °C, 80 °C and 120 °C at a depth x of 1.5 mm (top) and 5.0 mm (bottom).The images represent the situation in middle of the lms (y = 15 mm).Interfaces between layers are in some cases visible as thin dark horizontal lines.

Figure
FigureSI10presents POM images obtained at heights z of 0.6 mm and 10.2 mm.Similar to the POM images of PA12 thin lms, PLA thin lms show interfaces between the printed layers indicated by thin dark horizontal lines which become less pronounced with increasing bed temperature T b and height z.However, the interfaces are seemingly less extended compared as to PA12 and the components appear in general slightly more homogeneous.

FigureFigure
Figure SI 10: POM images (magnication 5k) of PLA thin lms with bed temperature Figure SI8 of 30 °C, 80 °C and 120 °C at heights z of 0.6 mm (top) and 10.2 mm (bottom).The images represent the situation in middle of the lms (y = 15 mm).

Table SI 1
: Terminal relaxation times τ R and shift factors a T for PA12 and PLA