The optical microscope included in all of the fabrication setups was used for sample imaging and analysis before and after development. The 50× oil immersion objective used in polymerization was also used for real-time monitoring of the polymerization process and for pre-development sample imaging. After the development procedure, a 20× air objective was utilized for imaging and visual evaluation of the sample quality. Scanning electron microscopy (SEM), atomic force microscopy (AFM), and wide-field fluorescence microscopy were used for more detailed characterization of the fabricated microstructures.
4.7.1 Scanning electron microscopy
In Publication I, PCL-o and PEGda structures were sputter coated with gold (S 150 Sputter Coater, Edwards Ltd., UK) in an argon atmosphere to a coating thickness of approximately 60 nm and imaged by SEM (Philips XL-30, Philips Electron Optics, the Netherlands). The dimensions of voxels, lines
46
and lattices were measured from the top view (0° tilt) or side view (25° to 30° tilt) SEM images with the free image processing software, GIMP 2.6 (GNU Image Manipulation Program, The GIMP Team). From the SEM images of the voxel arrays, complete, but still surface-bound voxels were selected for measurement purposes. As robust and measurable arrays of lines could only be fabricated from PEGda‐1.5 and PEGda‐1, simple lattice structures were used to study the combined effect of the average laser power and scanning speed on achievable feature sizes. The dimensions of the lattice walls were estimated from the side view SEM images.
In Publication II, protein samples were fixed in 5% (v/v) glutaraldehyde for 15 min, and then dehydrated by sequential immersion for 15 min in ion-exchanged water, 1:1 ethanol/H2O, 100% (v/v) ethanol, 1:1 ethanol/methanol and 100% (v/v) methanol (twice). After overnight drying in a desiccator, the samples were sputter coated with gold in an argon atmosphere to a nominal thickness of 75 nm or 15 nm (S 150 Sputter Coater or SCD 050 Sputter Coater, BAL-TEC AG, Liechtenstein).
The samples were imaged using either Philips XL-30 or JEOL JSM-6390 LV (JEOL Ltd., Japan) SEM. The line widths were measured from the top view (0° tilt) SEM images and the line heights from the side view (45° tilt) images with GIMP 2.6. The topography of the 2D and 3D protein patterns was also analyzed visually by comparing the SEM images of structures fabricated with the two different laser sources and different protein compositions.
Avidin and bBSA neuron guidance patterns crosslinked in the presence of Irgacure® 2959 (unpublished data) were sputter coated with gold to a nominal thickness of 38 nm (S 150 Sputter Coater) without any preceding fixation or dehydration steps. The samples were imaged with Zeiss ULTRA Plus FESEM (Carl Zeiss Microscopy GmbH, Germany) and the pattern dimensions were measured from the top view (0° tilt) and side view (60° tilt) SEM images with GIMP 2.8.
In Publication III, the fabrication accuracy of neurocages with the 2PP-DLW setup was evaluated by examining type III neurocages with SEM and by comparing the measured dimensions with the original CAD design. For SEM imaging, samples were sputter coated with gold in an argon atmosphere to a nominal thickness of 113 nm (S 150 Sputter Coater). Imaging was performed with a Philips XL-30 SEM. The acquired SEM images were analyzed with GIMP 2.8. The neurocage node diameter, channel length, channel width, and wall thickness were measured from the top view (0° tilt), and wall height from the side view (90° tilt) SEM images. The surface texture of the cages fabricated with the scanning speed of 120 µm/s was evaluated from the SEM images taken with a tilt angle of 60° by measuring the width of the ridges.
In Publication IV, the dimensions of the fabricated Ormocomp® suspended line structures and microtowers were analyzed by SEM imaging with Philips XL-30. Prior to imaging, the samples were sputter coated with gold in an argon atmosphere either to a nominal thickness of 113 nm (S 150 Sputter Coater) or 60 nm (SCD 050 Sputter Coater). The feature dimensions were measured from top (0° tilt) and side (90° tilt) view SEM images with GIMP 2.8. The degree of voxel overlap for the chosen contour distances was calculated as the product of voxel displacement in axial and lateral directions according to the following equation:
47 𝛿 = 𝑤 − 𝑑𝑥
𝑤 ×ℎ − 𝑑𝑧
ℎ , (13)
where δ is the degree of the voxel overlap, w and h are the width and height of the voxel, and dx and dz are the lateral and axial voxel distances, respectively (Žukauskas et al. 2012).
4.7.2 Atomic force microscopy
In Publication II, patterns of concentric squares of avidin and BSA were also studied by non-contact mode AFM (XE-100, Park Systems Inc., USA) before SEM imaging to verify the dimension measurements from SEM images and to study the surface topography of the structures. Measurements were performed using silicon probes (ACTa-910M, Applied NanoStructures Inc., USA) with a nominal resonance frequency of 300 kHz, spring constant of 4 N/m, a pyramidal-shaped tip (radius
< 10 nm) and an aluminum reflective coating. Images were acquired with a scan speed of 0.15 Hz or 0.40 Hz. The dimensions and surface topography were analyzed from AFM images with the XEI image processing software (Park Systems Inc., USA).
In Publication IV, the surface roughness of the microtower cylinder walls was investigated by non-contact mode AFM (XE-100). For the imaging, the coverslip containing the microtowers was mounted in upright position on the xy-scanning stage in order to access the surface of the walls of the towers with the AFM cantilever. Measurements were performed using silicon probes (ACTa) with a nominal resonance frequency of 300 kHz, spring constant of 40 N/m, a tetrahedral pyramidal-shaped tip with a face angle of 18°. Images were acquired with a scan rate of 1.0 Hz. Areas of 10 µm × 10 µm in size were imaged from the cylindrical walls of each tower design (IV, V, and VI). From the two designs with the openings, images were acquired from two different locations of the cylinder: the sections with smoother surface and the sections located between the openings with a seemingly rougher texture. The roughness of the surface was analyzed from AFM images with the XEI image processing software (version 1.8.0, Park Systems Inc., USA). The curvature resulting from the cylindrical shape of the tower wall was removed from the acquired images by utilizing the flattening tool with second order fitting curve in the x-direction. As the microtowers appeared to be tilted in the y-direction to some extent, the slope was eliminated by also flattening the images in the vertical direction with a first order-fitting curve. The surface roughness was defined as the average areal surface roughness (Ra).
The stiffness of the polymerized structures is mainly expressed by the Young’s modulus of the used material. It should be at a sufficient level for the microstructures to withstand the handling and cell culture procedures without failure. Thus, force spectroscopy measurements were performed with AFM to estimate the Young’s modulus of Ormocomp®. The elastic properties of the sample were investigated with AFM force curves by recording the applied force and the depth of indentation as the tip was pushed against the sample surface (Heinz & Hoh 1999). Force-indentation curves were recorded at three points from each tower design by scanning areas of 15 µm × 15 µm from the upper