Page 378 - IJB-10-4

International Journal of Bioprinting 3D cartilage induction and monitoring

min at 37ºC. After 14 days, the scaffolds were fixed with 4% The Multiphysics tool was employed to model the
paraformaldehyde (PFA) in PBS for 30 min at RT. Samples piezoelectric effect of the transducer to reproduce the
were treated with a primary antibody against collagen type transducer behavior. Due to patent protection, the
II (1:200; Abcam, UK), collagen type I (1:200; Abcam, materials employed for the piezo and the matching
UK), and aggrecan (1:200; Abcam, UK). Thereafter, they layer were unknown; after early trials, PZT-5H for
were incubated with AlexaFluor 645 (1:500; Abcam, UK). the piezo and Armco iron was selected due to their
Images were obtained using a Nikon Eclipse Ti microscope similarity. The transducer’s thickness was established as
((Nikon Instruments Inc., Japan) and analyzed with Image 13 mm, and its reference impedance to 50 Ω (Olympus
J software. v103-RM; Olympus, USA). For charge conservation,
the relative permittivity was selected as the dielectric
2.9. Scanning electron microscopy model:
Samples were incubated for 21 days (2 × 10 cells per
6
scaffold) and subsequently fixed with 2% glutaraldehyde
overnight at RT. The next day, they were rinsed in 0.1 D = є є E (III)
0
r
M cacodylate buffer and stored at 4°C. Samples were
then prepared for scanning electron microscopy (SEM) where є is the vacuum permittivity and є is the relative
0
r
according to the standard protocol: (i) several rounds of permittivity. No initial voltage was estimated, allowing the
washing with PBS; (ii) serial dehydration with ethanol following electric potential as input:
(30–100%); (iii) critical point drying in Emscope CPD
750 (EMZER Technological Solutions, S.L, Spain); (iv) Vt() = V ∗ sin(ω t)| T 0
mounting on aluminum SEM top stubs; and (v) sputter 0 0 0 (IV)
coating with conductive material (gold-palladium alloy,
Sputter Coater 108 Auto; Leica Microsystems INC, where V = 20 V and T = 1 µs. On the other hand, a
0
0
Germany). Finally, samples were imaged using an FEI linear elastic material with isotropic behavior was selected
Quanta 400 microscope (Thermo Fisher Scientific, USA) for the solid mechanics’ response of the transducer. A
with an Everhart-Thornley detector (E-TD) for dry and fixed restriction was imposed on the outer boundary of
conductive samples in high-vacuum mode. the piezo to prevent motion in that direction. A stress-
2.10. Statistical analysis charge form was applied to describe the constitutive
Three experiments were performed for each condition relation of the piezoelectric material. All parameters and
to assess variability (n = 3), and data representation geometries of the transducer were reconstructed using an
corresponds to mean ± standard deviation. Data were inverse problem with materials, such as water and PLA,
processed and presented using Origin 9.0 software whose speed of sound and density has been reported in
(OriginLab Corporation, USA) and were validated for the literature.
homoscedasticity (Shapiro-Wilk test). Student’s two-tailed Wave propagation was modeled with the Acoustic
t-test was used to analyze data with a confidence interval of Pressure tool, whereas the transitory model was linear
0.05. If present, outliers were ignored using the interquartile elastic, depending on the speed of sound and density
range (IQR) method, where multiple samples are compared of each material. A convergence study was conducted
using the Analysis of Variance (ANOVA) test, and means for ∆x (maximum mesh size for a triangular mesh) to
are compared using the two-tailed Bonferroni test with a ensure adequate wave propagation. The density and
confidence interval of 0.05. In the figures, *** denotes p < speed of sound parameters were obtained through the
0.001, ** denotes p < 0.01, and * denotes p < 0.05. inverse problem. The previous study demonstrated
the independence of parameters to mesh size. An
2.11. In silico model acoustic-structure boundary condition was used for
A computational model using Multiphysics FEM software boundary intersection. The model was performed in
was used to evaluate and understand the effect of the the time domain to obtain a comparable signal with
pressure wave exerted by ultrasound in the culture. For the experimental one; the time-dependent variable (t)
that purpose, a set of synthetic signals was obtained varied in a range from 0 to 50 µs (i.e., the same range as
through our FEM model, where several parameters were experimental time windows):
screened. In addition, shear stress induced by the laminar
flow was also modeled (COMSOL Multiphysics, USA). A
2D model was applied to reduce computational time and ∆t = T 0 (V)
memory consumption. The plane of the simulation was the 40
XZ middle plane of the BR.

Volume 10 Issue 4 (2024) 370 doi: 10.36922/ijb.3389