1 Introduction
Screen printing technique has been widely used in different industries, including clothing, product labels, signs, textile fabric, printed electronics and so on [1,2]. One of the parameters that can vary and can be fine controlled in screen printing is the thickness of the print films or patterns [3,4]. This makes it useful for some of the techniques of printing electronics etc. Particularly, the manufacturing of traditional silicon solar cells and emerging photovoltaics of mesoscopic solar cells involves the process based on screen printing [5–7].
The 1st generation mesoscopic solar cells of dye-sensitized solar cells (DSSCs) [8,9] are based on a sandwich structure of fluorine-doped tin oxide (FTO)/mesoporous TiO2/electrolyte/Platinum electrode/FTO. Since the TiO2 layer (10−12 mm) can be deposited with screen printing process [7,9], the manufacturing of DSSCs is easy to scale up [10–12]. Various large-area DSSC modules and panels based on screen printing fabrication process have been reported, providing promising prospect for practical applications [13–16]. The 2nd generation mesoscopic solar cells of perovskite solar cells (PSCs) employ organic-inorganic hybrid trihalide perovskites (typically CH3NH3PbI3) as the light absorbers [17], and usually possess a monolithic structure of FTO/electron transport layer (ETL)/perovskite/hole transport layer (HTL)/Au [18–20]. Currently, PSCs have achieved a certified power conversion efficiency (PCE) of 23.7% in laboratory-scale [21,22]. To further explore the potential commercialization, intensive research has been focused on the stability and upscaling of PSCs [23–27]. Depending on the sequence of depositing the ETL and HTL, the device structure can be divided into formal (conventional) versus inverted architectures [28,29]. Particularly, we have reported a triple mesoscopic structure based on the scaffold of TiO2/ZrO2/carbon for PSCs [32]. Since the deposition of the scaffold is entirely based on screen printing process, this type of PSCs is simply named as printable PSCs. Currently, this type of PSCs has demonstrated quite promising potential for the commercialization and attracted much research attention [33–36].
For fabricating efficient printable PSCs, the optimal thickness of the TiO2 layer, ZrO2 layer and carbon layer of the scaffold are ~500 nm, ~3 mm and 10 mm, respectively [37]. As usual, it is implementable for screen printing technique to deposit micrometer-thick films, such as the silver grids for silicon solar cell cells and TiO2 electrodes for DSSCs. However, to deposit nanometer-thick films by screen printing technique, it is challenging to precisely control the thickness and uniformity.
Here, we tune the concentration of the printing pastes and printing parameters for coating TiO2 films, and successfully print TiO2 films with the thickness of 500−550 nm and minimize the thickness errors. The correlation between the thickness of the films and printing parameters such as the solid content and viscosity of the pastes, the printing speed and pressure, and the temperature has been investigated. Besides, the edge effect that the edge of the TiO2 films possesses a much larger thickness and printing positional accuracy have been studied.
2 Experimental section
2.1 Materials
Unless stated otherwise, all materials were purchased from Sigma-Aldrich. The TiO2 was purchased from GreatCell Solar or WonderSolar. The ZrO2 and carbon pastes were provided by WonderSolar.
2.2 Printing of the films
FTO glass substrates were etched with a 1064 nm laser, and then ultrasonically cleaned with detergent, deionized water and ethanol for 10 min, respectively. The screen printer (SMT-DEK, icon 6) has an automatic sample loading system and a CCD automatic positioning system. The parameters of the screen mesh are as: 150T mesh/inch; tension 25 N/cm; polyester thread with a diameter of 45 mm. The printing parameters are as: printing speed 50–250 mm/s; printing gap: 2−4 mm; printing pressure: 4−10 kg. After printing, the films were sintered at 500°C for 30 min.
3 Results and discussion
The printable mesoscopic PSCs based on TiO2/ZrO2/carbon triple-layer were fabricated using screen printing techniques, as shown in Fig. 1. The TiO2 layer, ZrO2 layer and carbon layer were screen printed on the FTO substrate layer by layer. The perovskite absorber was infiltrated in the mesoporous scaffold by a simple solution casting method, as shown in Fig. 1(a). The detailed fabrication process can be found in the experimental section. To obtain high efficiency and reproducibility, it is essential to control the thickness of each layer of the scaffold. To fabricate large-area submodules, the relative positions of each layer is also important, which need to construct a series-connections for the strip unit cells, as shown in Fig. 1(b). The cross-sectional scanning electron microscopy (SEM) image of the screen printed scaffold is shown in Fig. 1(c).
For printing the TiO2 layer, the most widely used printing paste is supplied by GreatCell Solar (NR30). However, if the paste is used as received, the thickness of the printed TiO2 layer is usually 2−3 mm. Of course, the thickness can be tuned by optimizing the mesh and printing parameters, but it is impossible to reduce the thickness to below 1 mm. Thus, we diluted the TiO2 paste with the solvent of a-terpineol. Since the paste has a very high viscosity, it is more convenient to prepare the samples with mass ratio, not volume ratio. When we diluted the TiO2 paste with 3.5, 4.0 and 4.5 times of a-terpineol, the thickness of the printed TiO2 film significantly reduced to 725±16 nm, 636±16 nm and 559±13 nm, as shown in Fig. 2(a). The thickness was measured by a profilometer, and the results were summarized with 9 points on the samples. The statistical distributions of the thickness are shown in Fig. 2(b).
After diluting the TiO2 paste, the thickness of the screen printed TiO2 film has successfully reduced to the target value of 500−600 nm. To further study the factors that determine or mainly influence the thickness of the film, we measured the viscosity and solid content of the pastes. At room temperature (RT, 25°C), the TiO2 paste/terpineol 1:3.5 sample has a viscosity of 286.7 cP. When the ratio of terpineol increased to 4.0 and 4.5, the viscosity of the paste significantly reduced to 237.2 and 196.3 cP. This much reduced viscosity may influence the screen printing process and the film thickness. Particularly, the TiO2 paste is kept in the refrigerator (4−8°C) for storage. When it was taken out and used, the temperature usually cannot reach RT. Thus, we can observe that the viscosity of the paste varied when we printed it. Whether the viscosity affect the thickness of the printed films remains a critical issue.
We measured the viscosity of the TiO2 pastes with different terpineol ratio at different temperature, as shown in Fig. 3(b). It was found that the viscosity of the pastes is quite sensitive to the temperature. When the temperature increased from 18°C to 34°C, the viscosity of the pastes reduced from 400−600 cP to 100−150 cP. In the clean room, the temperature is set to 25°C, but there might be a temperature error of at least 2−3°C. In this case, the viscosity of the paste may have a deviation of 10%−20%. To verify whether the viscosity affect the film thickness, we printed TiO2 films with 1:4.5 paste at different temperatures, as shown in Fig. 3(c). It was found that the film thickness slightly increased 518 to 642 nm when the viscosity of the paste reduced from 397.70 to 80.21 cP. If we can control the temperature variation to be 2−3°C, the variation of the viscosity of the pastes will be<100 cP. In this case, the variation of the film thickness will be less than 50 nm. In another word, the viscosity of the paste and the printing temperature may influence the film thickness, but are not the main factors. The printed film thickness is determined by the amount of the materials that can go through the screen mesh and transfer to the substrate during the printing process [4]. The lower the viscosity of the paste is, the easier the paste can pass through the screen mesh and transfer to the substrate. Thus, reducing the viscosity of the paste can increase the thickness of the printed films.
More directly, the film thickness is influenced by the solid content of the pastes, as shown in Fig. 3(d). This is the key why the film thickness can dramatically reduce when the pastes were diluted. It is not caused by the reduced viscosity, but the materials than can be transferred to the substrate with the same screen mesh and printing process. The paste may pass through less, but the final solid-state materials on the substrate is much more, leading to increase film thickness.
As mentioned previously, changing the screen mesh with different parameters, such as the mesh count, tension and thread diameter, can also tune the film thickness. Here in this work, the mesh count of the screen mesh is 150T mesh/inch. The tension is 25 N/cm, and the diameter of the polyester thread is 45 mm. Besides the screen mesh, the printing parameters also play a key role in controlling the film thickness.
As shown in Fig. 4, the influence of the main printing parameters of print speed, print gap and print pressure on film thickness was investigated. When the print speed decreased from 250 to 50 mm/s, the film thickness reduced from 725±16 to 500±23 nm. As mentioned previously, the film thickness is determined by the amount of the paste that can go through the screen mesh. Here the amount of the paste that can go through the screen mesh is influenced by the deformation of the screen mesh. When the squeegee moves with higher speed, the screen mesh will undergo a more severe deformation, leading to more paste go through the screen mesh and transfer to the substrate. Thus, the higher printing speed will increase the TiO2 film thickness. The print gap is the distance between the screen mesh and the substrate. When it increased from 2 to 4 mm, the film thickness increased from 649±13 to 757±23 nm. For the print pressure, there is an optimal value for the paste to pass through, obtaining the thickest films. When the pressure is too low, the screen mesh may not be able to contact the substrate tightly, slowing down the transfer of the paste. When the pressure is too high, the screen mesh is squeezed too much, which also suppresses the transfer of the paste from the mesh to the substrate.
Furthermore, we investigated the printing accuracy of the printed TiO2 films. The screen printer is supplied by SMT-DEK (icon 6), and has a CCD automatic positioning system. For the pastes with different viscosity and solid content, the printing accuracy may vary from 88±8 to 98±9 nm, as shown in Fig. 5(a). We etched the FTO layer on the FTO substrate, forming a target position, as shown in Fig. 5(b). With a microscopy system, it was clear observed that the printed TiO2 films have quite rough edges. Besides the positional accuracy, we also investigated the printing length accuracy, as shown in Fig. 5(c), which is much smaller than the positional accuracy. This indicates that printing error is not only caused by the spreading of the pastes on the substrate after printing, but also by the printing accuracy of the printer.
Another issue is the edge effect of the printed TiO2 films, which is widely observed for screen printing technique [38,39]. As shown in Fig. 5(d), the edge of the film (Paste 1, GreatCell Solar, NR30) has an extremely high peak with a thickness of over 1500 nm, while the target thickness of the films is only 500−550 nm. For the sample prepared with paste 2 (purchased from WonderSolar, China), the edge effect is avoided. Notably, the paste 2 is used as received without long-term storage. The exact causes of the edge-peaks have not been ween identified, but it is highly possible to be related to the aggregation effect of the paste after a long-term storage. Maybe fresh paste 1 can also obtain edge-peak-free TiO2 films. Now that the edge effect can be effectively avoided, the potential further large-scale production of the such TiO2 films and printable PSCs be much more practical.
4 Conclusions
We tuned the printing pastes and printing parameters for coating TiO2 films, and successfully print TiO2 films with the thickness of 500−550 nm. The influence of the solid content and viscosity of the pastes, the temperature, the printing speed, gap and pressure on the film thickness was investigated. It was found that the thickness was determined by the fact that how much paste can be transferred from the screen mesh to the substrate. The edge effect that the edge of the TiO2 films possesses a much larger thickness and printing positional accuracy have also been studied. This work will significantly benefit the further development of printable mesoscopic PSCs.