Thin film deposition process
Aluminum layers of 2019 Å over microscope slides were obtained by PVD, Fig. 1a shows a sketch of this technique. After the cleaning process described in section 2.3, glasses were introduced in the coating system where the aluminum layer was deposited at a deposition rate of 50 Å/s under a vacuum of 2 × 10− 5 mbar.
After that, the metalized substrate undergoes an annealing process at 200 °C for 2 h in an oven, that notably enhances the quality of the soldering afterwards and homogenize the coating surface [28]. Images of the profile of the layer have been taken by measuring the cross section of the samples with Scanning Electron Microscope SEM (Fig. 1b). Before the annealing processes, the layer presents a roughened surface with frequents clusters of particles. However, the thermal annealing leads to a smoother surface. In this case, the image was taken with a working distance of 5.2 mm, an extra high tension (EHT) of 15 KV and 8 × 104 magnification.
Electric conductivity of the samples was measured before and after the thermal treatment obtaining that before the thermal treatment the electric conductivity of the layer is 2.250 ± 0.085 × 105 S/cm while after the annealing process, the conductivity is 2.492 ± 0.086 × 105 S/cm. Taking into account that the conductivity of the aluminum is 3.77 × 105 S/cm, it can be asserted that it is reduced in a 40% due to the deposition process. However, it can be increased carrying out the annealing process.
Laser fabrication of electrical tracks
The fabrication of the electrical stimulator circuit is based on the laser ablation technique. Aluminum layers were irradiated with a quasi-perpendicular nanosecond Nd:YVO4 laser emitting at a wavelength of 1064 nm. The laser is fitted with a galvanometer beam steering system and a flat-field lens of 100 mm focal distance (Fig. 1.d). This lens allows scanning the substrate within the X, Y plane. In this technique, the laser is focused onto the surface of aluminum layer generating an intense plasma plume that pull off Al ions and particles from the layer without causing any cracks on substrate (glass) surface. A specific pattern structure, generated by CAD-like software, was designed, and fabricated selectively removing part of the aluminum layers deposited as described in section 3.1.
Figure 1d, shows the CAD design used in electrical stimulator fabrication. Laser parameters setting were optimized in order to perform the metal layer removal without damaging the substrate. An average laser power of 1.05 W, a repetition rate of 12 kHz and a scan speed of 60 mm/s were selected to perform the ablation process. The aluminum was successfully removed using a laser fluence value between the damage threshold of glass and the ablation threshold of the target (920 J/cm2 and 4.20 J/cm2, respectively).
To determine the threshold fluence value for the aluminum layer, we follow the method of Liu [35].. Assuming an output gaussian beam for the laser, the spatial fluence (φ(r)) is given by
$$ \varphi (r)={\varphi}_0{e}^{\raisebox{1ex}{$-2{r}^2$}\!\left/ \!\raisebox{-1ex}{${\omega}_0^2$}\right.} $$
(1)
where ω0 is the gaussian beam radius (measured at 1/e2), φ0 is the peak fluence of the laser and r is the distance from the center of the beam. The energy per pulse (Epp) and the peak fluence are related according to the equation
$$ {\varphi}_0=\frac{2{E}_{pp}}{\pi {\omega}_0^2} $$
(2)
The relation between peak fluence and the diameter of the laser ablated spot (D) is given by
$$ {D}^2=2{\omega}_0^2\mathit{\ln}\left(\frac{\varphi_0}{\varphi_{th}}\right) $$
(3)
Combining eqs. 2 and 3 following relation is obtained
$$ {D}^2=2{\omega}_0^2\mathit{\ln}\left({E}_{pp}\right)-2{\omega}_0^2\mathit{\ln}\left(\frac{2}{\pi {\omega}_0^2{\varphi}_{th}}\right) $$
(4)
Therefore, using above equation threshold fluence and Gaussian beam spot size can be obtained by measuring the diameters of the ablated areas D. Plotting D2 versus the ln (Epp) and fitting the data, ω0 is determined and the threshold fluence is obtained extrapolating to D2 = 0. Results are shown in Fig. 2a, where values of 4.19 ± 0.77 J/cm2 and 25.69 ± 0.51 μm were obtained for aluminum layer threshold fluence and for the radius of the beam, respectively.
The proper selection of the laser parameters allows us to ablate the aluminum in the most efficient way without causing damage in the glass substrate. The adjustable parameters of the system used are the laser fluence, the laser frequency and the scan speed of the beam. To determinate the ratio between the laser frequency and the scan speed we define the degree of pulse overlap between consecutive spots. This factor is set in the equation
$$ {O}_d=1-\frac{v}{2 df} $$
(5)
where v and f are the scan speed and frequency, respectively, and d is the diameter of the spot crater. Pulse overlapping is a crucial parameter for fabricating a homogeneous electrical track. Excessive overlap will deliver too much energy to the glass, damaging the surface and increasing manufacturing time, while low overlap will result in inefficient material removal.
In Fig. 2b electrical tracks of a 200 nm aluminum layer ablated with different laser frequencies and scan speeds and therefore, with different pulse overlapping are presented. All of them were ablated with an average power of 700 mW. Results show how the ratio of 12 kHz and 60 mm/s produces tracks with more regular width. This correspond to an overlap degree of 0.66. In other cases, the pulses are either too separated or too overlapped. Based on these results, electrical tracks were fabricated using an overlapping factor of 0.66.
Finally, the optimal ratio between laser power and frequency was also analyzed. Tracks were fabricated with different values of both parameters. Figure 2c shows the diameter of a single line versus the energy per pulse, obtained by using different combinations of frequency (from 8 to 18 kHz) and power (from 700 to 2000 mW). It can be observed the linear relationship between the width of the line and the energy per pulse. This width was considered during the aluminum layer removal process. Therefore, the optimum laser parameters setting for the electrostimulator fabrication were, a pulse energy of 90 μJ (corresponding to a frequency of 12 kHz and a laser power of 1.05 W) combined with a scan speed of 60 mm/s.
In Fig. 2d, it can be observed an optical microscope image of a part of the electrical stimulator. The microscope is equipped with a double lighting system that allows illuminating samples both from above and below. In the Fig. 2d, orange region corresponds to the glass substrate and the green one to the aluminum tracks. Note that the image was taken lighting the electrical stimulator from below. The aluminum was successfully removed and there are no contacts between the tracks. Besides, the glass substrate was not damaged.
Electrical tracks characterization
To study the shape and intensity of the electrical field generated by the circuit, simulations were performed using the software ANSYS Maxwell. Simulations were performed applying a voltage of 5 V between the terminals of different models. In each case, results show an electrical field fluctuating spatially between two values, with a period equal to the distance between the center of two consecutive tracks. This behavior can be observed in Fig. 3 a and b.
Figure 3a shows the intensity of the electric field induced on the surface of a 145 μm glass (microscope cover slip) by an aluminum circuit placed underneath. In this figure, a circuit with track width of 225 μm and a gap of 50 μm between tracks was simulated. A cross section of the electrical field propagating through the thin glass is shown in Fig. 3b. Note the oscillatory behavior of the electrical field profile at the interface glass-air in the orthogonal direction to the tracks. As consequence of this behavior, it is necessary to define the homogeneity of the electrical field as
$$ H=\frac{{\left|\overrightarrow{E}\right|}_{min}}{{\left|\overrightarrow{E}\right|}_{max}} $$
(6)
According eq. (6), the electric field will be more homogeneous when the value of H approaches 1, while the oscillations will be more relevant the closer the value of h approaches zero. The electrical field induced by the circuit along the direction perpendicular to the circuit tracks (represented with a white line in Fig. 3a) was characterized for tracks widths in a range from 20 to 5000 μm and for gaps between them from 20 to 500 μm. Results of the peak value of the electrical field are shown in Fig. 3c, and Fig. 3d shows the homogeneity of the electrical field for the same circuit geometries. Simulations show that electrical field amplitude increases with the width of the tracks while the homogeneity of the field over the cover slip decreases. On the other hand, in Fig. 3c two regions are clearly differentiated. If the tracks width is less than 300 μm, those circuits with a greater gap between tracks will provide a greater field. On the contrary, when the width of the tracks exceeds 300 μm, the circuit that produces the largest electric field will be the one with the smallest gap between tracks. These two regions are also distinguished in Fig. 3d where it is shown that more homogeneous electric fields are generated for those circuits with less gap between tracks. Furthermore, the homogeneity is almost constant for tracks width less than 200 μm and reaches a maximum depending on the gap between the tracks. An abruptly decrease of the field homogeneity is observed for tracks width between 200 and 500 μm and gaps from 200 to 500 μm.
To study electrical cell stimulation, a low homogeneity electric field was selected. Figure 3e shows the electric field corresponding to a circuit with 3 mm width tracks and 20 μm gap that provides an oscillating electric field between a minimum of 455 V/m and a maximum of 2437 V/m.
Once the manufacturing process was completed, the programming of a control software that allowed to apply different electrical signals to cell cultures was addressed. With this purpose a specific program connected to an I/O device which generate square signals according to the user’s indications was developed.
The hardware selected to provide the signal was the NI USB-6501 portable digital I/O device, from National Instruments. It provides 5 V by default and up to 8.5 mA. For this application, the program was designed to apply a non-symmetrical square waveform with an amplitude of 5 V. The frequency of the signal, the duration of the pulse and the duration of the signal can be tuned by the user. Almost any modern computer is enough to run and manage both the hardware, which only needs a free USB slot, and the software, which no requires having LabVIEW installed.
Cell stimulation
Cell Imaging Dishes (Eppendorf, Hamburg, Germany) containing H9c2 cells (50,000 cells were seeded in each cell dish) was placed directly over the electrical stimulator circuit (Fig. 4a and b) and a continuous signal of 5 V was applied, inducing over the surface of the cell imaging dishes the electrical field showed in Fig. 3e. The electrical stimulus (5 V), was applied throughout the experiment for 24, 48 and 72 h. After the experiment, cells were marked with the DAPI fluorescence stain. Fluorescence microscopy images were taken using the Leica TSC SP8 confocal microscope. Cell density was determined by processing the images with the software ImageJ (minimum particle size of 50 pixels and minimum circularity of 0.3).
Figure 4c shows a fluorescence microscopy image (the blue dots are the cell nucleus). Figure 4d shows the same image after being processed by the ImageJ software, which identifies the nucleus for an accurate count. To evaluate the number of cells fluorescence microscopy images with an area of 1.385 mm2 were analyzed (between 4 and 10 depending on the quality of the image for each condition). Results are presented as the mean value of all the images processed and the error as 2σ (the half-width of a 95% confidence interval). Results of the cell density count are shown in Fig. 4e for the samples exposed for 24, 48 and 72 h, as well as the count in the control samples. It can be observed a remarkable increase in the number of cells for the stimulated samples, where after 72 h the cell density doubles the cell density of the control samples. Note that the bovine serum was removed from the culture medium to avoid any effect on the cells so that cells proliferation is due solely to the effect of the electrostimulation.