Page 321 - IJB-9-1
P. 321

International Journal of Bioprinting                                   Sterilization effect on 3D printed inks



            Table 1. The parameters of Cross model for GEL-SA samples as a function of the sterilization method (CONTROL, UV, FILT, and
            AUTO) and media (DMEM, PBS, and water) used.
             Sample                η 0            η ∞            C              m              R 2
                                   (Pa·s)         (Pa·s)         (s)
             DMEM-CONTROL          446.79         0.00           0.617          1.244          0.99
             DMEM-UV               255.72         0.37           0.619          1.277          0.99
             DMEM-FILT               3.70         0.60           0.153          1.611          0.99
             DMEM-AUTO               0.42         0.16           0.555          0.617          1.00
             PBS-CONTROL           412.19         0.00           0.321          1.468          0.99
             PBS-UV                268.80         0.45           0.430          1.299          0.99
             PBS-FILT                1.81         0.32           0.178          1.186          1.00
             PBS-AUTO                0.18         0.13           0.197          1.639          1.00
             Water-CONTROL         629.57         0.00           0.377          1.406          0.99
             Water-UV              295.98         0.64           0.297          1.475          0.99
             Water-FILT              0.17         0.10           0.075          0.773          1.00
             Water-AUTO              0.10         0.07           0.104          0.471          1.00
            CONTROL: non-sterilized inks; UV: UV-sterilized inks; FILT: filtered inks; AUTO: autoclaved inks

































            Figure 3. Storage modulus (G'), loss modulus (G") and loss tangent (tanδ) as a function of frequency for: (A) non-sterilized (CONTROL), (B) UV-
            sterilized (UV), (C) filtered (FILT), and (D) autoclaved (AUTO) GEL-SA samples in DMEM media.


            set out in Table 1. All samples, except AUTO and water-  (C) and Cross rate constant (m) values, which determine
                                                                                      ⋅
            FILT samples, demonstrated shear thinning behavior,   the shear rate in the nozzle ( γ   ), are also listed in Table 1.
                                                                                       w
            as evidenced by the difference between µ  and µ values,
                                              0
                                                    ∞
            which showed a drop in viscosity, especially for the   In addition, frequency sweeps were carried out to
            DMEM-CONTROL sample. This shear thinning behavior   determine storage (G’) and loss (G”) moduli, as well as
            is desirable for 3D-printing applications since viscosity   loss tangent (tanδ), and the resulting curves are shown in
            will decrease when shear rate increases during extrusion   Figures 3–5. As can be seen, non-sterilized and UV-sterilized
            and thus, sample could be extruded [44–46] . In contrast,   samples (graphs a and b) showed elastic behavior since the
            no  relevant  difference  was  found for  AUTO samples  in   values of G’ are higher than those of G”, regardless of the
            DMEM, PBS, and water media. The Cross time constant   media used; in contrast, FILT and AUTO samples showed


            Volume 9 Issue 1 (2023)                        313                      https://doi.org/10.18063/ijb.v9i1.645
   316   317   318   319   320   321   322   323   324   325   326