The use of nanostructured materials is increasingly widespread thanks to their particular properties that can improve the performance of devices in various scientific applications. One of them is in the architecture of perovskite solar cells characterized by high photoconversion efficiency values that make them able to compete with silicon solar cells. In this framework, we deposited TiO sponges by reactive sputtering based on a grazing-incidence geometry combined with the local oxidation of species. The deposited material gains 50% porosity in volume through depths of hundreds of nanometers and consists of a forest of uniform rods separated by mesopores (pipelines) arising from the grazing geometry. Many previous studies showed how TiO can improve the efficiency of perovskite solar cells. In this article, we investigated the change of the wettability values of the TiO samples before and after a postdeposition thermal annealing treatment. For comparison, the influence of the annealing on the wettability of the glass substrate is also reported.
I. INTRODUCTION
Perovskites are worldwide highly investigated materials due to their unique properties impacting various fields, such as photovoltaic,1 Light Emission Diodes (LEDs), and photodetectors.2,3
Perovskite solar cells have reached the exceptional goal of 25.5% of efficiency just ten years after the first publication by Miyasaka and co-workers in 2009 (Fig. 1).4 This result is comparable to what is achieved by traditional silicon cells and has further advantages, such as the wide absorption range,5 low production costs, and versatile processes.6
All their exceptional properties, which benefit from the possibility of preparing high-quality films in terms of crystallinity, thickness, and low defect density, are strictly connected with each other. These advantages are counterbalanced by issues that are still open. They are related to the device instability particularly under humid air ambient7 and to the current realization of large area films via wet deposition approaches, which cool down their attractiveness from the industrial point of view. A great effort has been dedicated to improving the stability and crystallinity of perovskite polycrystalline films.8 Dye-sensitized solar cells (DSSCs) have appeared as a promising source to meet the energy demand as an environmentally friendly alternative.
To boost the material and device performance, one approach is the use, implementation, and tailoring of highly performing nanomaterials. Among them, titanium dioxide (TiO), since its discovery for water photolysis by Fujishima and Honda in 1972,9 has progressively raised interest and has consequently been used in many fields, such as photocatalytic degradation of pollutants, photocatalytic CO reduction into energy fuels, water splitting, solar cells, supercapacitors, biomedical devices, and lithium-ion batteries.10 In the photovoltaic field, recent studies linked to the application of porous TiO in hybrid solar cell architecture led to increased efficiency.11
In Fig. 2 is shown a typical perovskite-structured solar cell architecture where the TiO layer plays the role of the electron transport layer. In the standard scheme of a photoanode,12 a mesoporous layer of nanosized TiO crystals is deposited on a transparent conductive oxide, annealed for grain sintering plus anatase crystallization (typically at 450 C) and subsequently soaked with a photoactive dye or by a perovskite material for light harvesting. However, the role played by TiO in constructing perovskite solar cells is strongly affected by different preparation methods (high/low temperature) and various optimization parameters (modulation of structure, size, shape, and interface interaction).
The use of chemical methods is the widespread and versatile way to generate nano-TiO architecture with mesoporous or hierarchical structures with high infiltration capability.13 Several attempts were done in the literature in order to grow porous TiO layers by physical approaches, mainly applying a model proposed by Thornton.14 To overcome the porosity limit of this approach, cauliflower,15 penniform,16 or zig-zag17 structures were also created. Instead, a different approach in the context of physical deposition solutions, concerning the advent of modified sputtering methodologies to grow highly porous TiO scaffolds, has opening new ideas for possible future applications.18 Recently, glancing angle deposition methods19 were used to deposit Ti nanostructures20,21 by ex situ oxidation; templating materials (e.g., polystyrene nanospheres) were applied to exploit their shadowing effect during TiO growth.22
Here, we report on designed local oxidation coupled with grazing-incidence sputtering methodologies used to grow multiporosity TiO layers. It consists of an industrially implementable approach that combines the reproducibility of the materials with the upscalability of the deposition process10 extending the possibility of depositing metal oxides on different substrates. The spongy layer provides as the main benefit of bimodal porosity, with a finely interconnected matrix of pores of convenient size for dye infiltration and anchoring. The high level of porosity and the fine structure of the material allow, for example, N-719 molecules to deeply and copiously enter the layer and establish a multibranch anchoring on the pore walls.10 Further main advantages of TiO gig-lox synthesized films with respect to a commercial TiO are reported in Table I.
Commercial porous TiO2 . | Gig-lox . |
---|---|
Solvent | Solvent-free |
Solvent release needed | No solvent release |
Deposition only on some | Deposition on |
specific materials | any materials |
Annealing ∼500 C° | Optional annealing ∼450 C° |
∼50% porosity in volume | ∼50% porosity in volume |
Commercial porous TiO2 . | Gig-lox . |
---|---|
Solvent | Solvent-free |
Solvent release needed | No solvent release |
Deposition only on some | Deposition on |
specific materials | any materials |
Annealing ∼500 C° | Optional annealing ∼450 C° |
∼50% porosity in volume | ∼50% porosity in volume |
In order to further optimize the surface properties of our TiO sputtered films, in this work, we report preliminary results about the modification of TiO film’s surface properties in terms of adhesion and wettability. These latter properties are fundamental to improving solar cells performance. Usually, during the different stages of solar cell manufacturing, contact angle measurements are performed to achieve useful information on the surface properties. In particular, the higher the water contact angle, the lower the water adsorption by the porous structures on the surface. Furthermore, superhydrophilic surfaces having a water contact angle close to or even less than 5 allow a reduction of possible contamination, leading to self-cleaning behavior.24 In addition, a contact angle measurement is used as an empirical diagnostic method to pre-evaluate the performance of dye-sensitized solar cells. The effect of dye-loading time on the surface wettability and solar-to-electrical conversion performance of dye-sensitized solar cells is undertaken.25
To this purpose, the static (equilibrium) contact angle values at room temperature (RT) were estimated. Different measurements were carried out at different points on each sample. Furthermore, the influence of a postdeposition annealing treatment was also analyzed since the annealing would affect wettability and film crystallization of the deposited TiO and the corresponding performance of the perovskite solar cells.
II. MATERIALS AND METHODS
A. TiO2 deposition
Spongy TiO layers were grown by an approach, developed at CNR-IMM, based on grazing-incidence sputtering deposition of titanium species from a metallic source (Sematrade supplier; 99.99% purity) coupled with the progressive local oxidation of the generated species at the growing front. This growth method26 was improved by optimizing the angular configuration to obtain an increase of the porosity. The experimental setup is reported in Fig. 3.
Our approach enables the following actions: (1) a high deposition rate (4 nm/min for 100 min) due to the metallic plasma established at the source side (charging effects by surface oxidation are indeed avoided); (2) the progressive local oxidation (monitoring the flow of oxygen and Ar during the process) of the landing species at the anode side; (3) the shadowing effect from the starting seeds due to the inclined (off-axis) titanium flux. These factors provide the material with additional mesoporosity (besides the nanoporosity arising from Thornton’s approach18) that is maintained during (eventually needed) postdeposition thermal treatments and up to 1000 nm.
We use ultrahigh purity (99.99%) gas sources for Ar and O. The base pressure that we achieved prior to introducing the gas species was mbar obtained by a turbomolecular pump. This vacuum level allows depositing at RT and obtaining uniform and contaminant-free films.27 Figure 4 displays an example of TiO gig-lox deposited on a perovskite solar cell after the sputtering process.
The deposited sample was annealed at 450C for 30 min since it is well known that mesoporous TiO in the anatase polymorphism has emerged as the most appropriate choice for cell scaffold.26,28,29
B. Characterization techniques
Some different characterization techniques were adopted to study the samples of chemical-physical properties. UV-Vis optical absorption measurements were recorded in the visible range, with a nm resolution, using a Perkin-Elmer Lambda 2 spectrophotometer. Micro-Raman measurements were performed using a Horiba XploRA spectrometer equipped with a confocal microscope and a Peltier-cooled charge-coupled detector. The samples were excited using the 532 nm line from a solid-state laser. X-ray Photoelectron Spectroscopy (XPS) was carried out in an ultrahigh vacuum condition using a Thermo Scientific instrument equipped with a monochromatic Al-K alpha source (1486.6 eV) and a hemispherical analyzer. The constant-pass energy was set at 200 eV for survey scans. Spectroscopic ellipsometry (SE) data were collected using a J. A. Woollam VASE instrument in a vertical configuration, which is better suited for transparent samples in order to measure on the same point ellipsometric and transmittance data. Optical spectra were recorded from 300 to 2100 nm (step 5 nm). An initial model of the optical transitions was built for each layer constituting the sample. The TiO layer was modeled using a single Tauc–Lorentz oscillator and the surface roughness using the Bruggeman effective medium approximation.30–32 The technique was mainly used to measure the average porosity of the layer. Then, Transmission Electron Microscopy (TEM) images were acquired using a Jeol JEM 2010 microscope in a diffraction contrast mode.
Contact angle measurements were carried out using the experimental homemade setup schematized in Fig. 5. It consists of an LED lamp of 580 lm at 6500 K, a USB uEye camera model UI-146xLE-C equipped with a 1/2 CMOS sensor, and an objective (lens) with a focal length of 16 mm and a max relative aperture of 1:1.4. A small drop of distilled water (2 l) is gently dropped onto the sample by means of a micrometer-controlled microsyringe. The small drop volume (about one-fifth of the water capillary length) and the small distance between the needle tip and the stage (less than 1 cm) allow us to neglect gravity effects. We have waited much time after sample annealing (3 days) to avoid temperature effects on the spread of the drops. The high-definition camera is placed on the same level to take a photo of the drop, which is then analyzed by a custom script written in MATLAB® to determine the value of the contact angle and the corresponding surface energy.
C. Contact angle theory
One of the most important variables to be measured for achieving information on the surface properties of a substrate is the water contact angle. In fact, the spreading extent of a water drop deposited on a solid substrate strongly depends on the interface interactions taking place between water and the substrate.33 In particular, the intensity of intermolecular interactions at the interface determines the equilibrium conditions and the corresponding drop shape.34 In terms of surface tension, the mechanical equilibrium is governed by Young’s equation,
where is the contact angle, and the three terms correspond to the interactions exerted by the surface tension at the three interfaces, which are solid-vapor (), liquid-solid (), and liquid-vapor (), schematized in Fig. 6. Note that the surface tension corresponds to the surface energy density so that Eq. (1) can be rewritten as
where is the surface energy of the solid substrate, is the interfacial tension between the liquid and the substrate, and is the surface tension of the liquid. Then, by invoking concepts related to the work of adhesion (), corresponding to the energy released in wetting (i.e., the work needed for separating the two phases), which is related to the surface energy of the two phases by the Duprà’s equation,
Then, the interactions between the liquid and solid, expressed in terms of the work of adhesion, are related at first approximation to the liquid surface energy and the contact angle. Therefore, by evaluating the water contact angle, an estimation of can be obtained. The contact angle value provides an evaluation of the wettability of the sample: for contact angle values above 90, the solid has poor wetting and is termed hydrophobic, whereas for values below 90, the sample is hydrophilic. However, there are many theories that try to take into account all the different contributions, such as dispersion and hydrogen bonding, participating in wetting phenomena.35,36
III. RESULTS AND DISCUSSION
Figure 7 shows a photo of TiO films deposited by the gig-lox method over substrates of different nature. In detail, Fig. 7 shows TiO layers (400 nm thick) deposited on (a) PEN (polyethylene naphthalate, a flexible and semitransparent substrate), (b) glass (fully transparent), and (c) SiO. All surfaces were cleaned with isopropanol before the deposition.
One of the strengths of the gig-lox method is the possibility to deposit the material potentially on substrates with different chemical/physical and mechanical characteristics that is useful for many applications.26 This particularity is due to the fact that the deposition takes place at low temperature and low pressure in the chamber, as reported in Ref. 26.
The chemical-physical properties of these films have been largely studied by our research group in the last few years as reported in Refs. 18, 26, and 37 and partially reported below. However, in this paper, we focalized attention on the hydrophilicity of TiO surfaces, deposited on glass, which can be suitably tailored via thermal annealing.
Figure 8 shows Raman spectra related to a 400 nm-thick TiO layer before and after the thermal treatment at 450 C for 30 min in air. In the lower-frequency region of the spectra, the features arising from titania vibration modes were detected.38 The comparison with reference library spectra39 reveals that, before annealing, the material is a mixture of anatase and rutile-TiO (short-range ordering). After thermal treatment at 450 C, the TiO structure gains a long-range order in the anatase atomic arrangement. Raman evidence has been confirmed by XRD analyses.12
According to literature data40 and from Raman evidence, we will expect that the anatase film will show long-lasting hydrophilicity. The effect of the material physicochemical properties on the wettability will be better discussed below.
The optical transmittance spectra of the TiO films were recorded in the range of 300–900 nm. Both samples are transparent up to about 350 nm. The oscillations in transmittance at a wavelength higher than 375 nm are ascribed to the interference due to the film thickness (about 260 nm) (Fig. 9).
Chemical environment of the component species in TiO films was investigated by XPS. XPS survey spectra are shown in Fig. 10. Both spectra display, together with the carbon contribution due to residual preparation, Ti 2p3/2 and Ti 2p1/2 spin–orbit doublet features located at 459.3 and 465.0 eV, respectively. This indicates that the oxide surface consists of nearly stoichiometric TiO.
The roughness degree of TiO films is an important parameter for the performance of solar cells. The surface roughness values, estimated by ellipsometric measurements, are 12.88 and 3.59 nm for the as-deposited and thermal annealed samples, respectively. We outline that a forest-like array of rod structures characterizes the annealed samples [see SEM images in Fig. 4(a) of Ref. 10]. The porous structure is more evident looking at the TEM cross-sectional image [Fig. 4(c) of Ref. 10]. Similar features are observed for the as-deposited film. Here, we show in Fig. 11 the TEM cross-sectional image of the as-deposited film in which we distinguish both nanopores and mesopores with 50% porosity in volume. More details are reported in Ref. 18. To increase the efficiency of DSSCs, mesoporous layers with high infiltration capability are generally required.11 We have demonstrated in a previous study (Ref. 10) that the interconnected array of pores allows deep- pervasive infiltration and multibranch chemisorption of the N-719 dye molecules through the whole layer. Moreover, we have measured an increment of the electron density in the conduction band of the gig-lox TiO layer of 4 orders of magnitude under 1 sun illumination with respect to the intrinsic carrier level. This indicates good injection efficiency, related to the dye density, distribution, and anchoring, and also on the carrier collection capability of the gig-lox TiO owing to the observed percolation path for the current to cross the whole material.
In recent years, there has been increasing interest in the wettability of oxide semiconductors as smart materials;41 however, the wettability of TiO has not been fully studied, especially its wettability conversion, also in relation with the sample morphology characteristics.
Figure 12 shows the images of the water drop on: (a) glass, (b) annealed glass, (c) as-deposited TiO film, and (d) annealed TiO film. The contact angle values estimated from these images are reported in Table II. A significant decrease in the CA value going from the as-deposited TiO sample (97.2) to the annealed one (12.9) is observed. We also take into account the CA changes of the glass substrate at RT (35.7) and of the glass annealed at T = 450 C (22.5). These outcomes indicate that the annealing treatment strongly influences the glass substrate wettability and mainly that of TiO films.
Sample . | CA (°) . | Wsl (mJ/m2) . |
---|---|---|
Glass substrate | 35.7 ± 1.3 | 131.92 ± 0.15 |
Glass substrate after annealing | 22.5 ± 1.5 | 140.06 ± 0.16 |
TiO2 as-deposited | 97.2 ± 1.2 | 63.68 ± 0.13 |
TiO2 after annealing | 12.9 ± 1.5 | 143.76 ± 0.16 |
Sample . | CA (°) . | Wsl (mJ/m2) . |
---|---|---|
Glass substrate | 35.7 ± 1.3 | 131.92 ± 0.15 |
Glass substrate after annealing | 22.5 ± 1.5 | 140.06 ± 0.16 |
TiO2 as-deposited | 97.2 ± 1.2 | 63.68 ± 0.13 |
TiO2 after annealing | 12.9 ± 1.5 | 143.76 ± 0.16 |
Note that the annealed sample has the lowest value of the water contact angle, being about 12, so showing an almost superhydrophilic character. Indeed, it has the highest value of the work of adhesion confirming that the droplet can spread more on the annealed surface. This means that water drops can easily form a film on the surface able to take away eventual contaminating dust.
Structural characterizations suggest that in addition to roughness, TiO rutile and anatase film phases affect hydrophobic and hydrophilic sites, which might be inducing specific wetting character.42 The higher hydrophobicity shown by the as-deposited film could be ascribed to its amorphous phase (i.e., lattice defects), while thermal annealing treatment induces higher film crystallinity with a higher hydrophilic degree.43
Although preliminary, the present results are, in our opinion, interesting. We have already planned the employment of other solvents, such as diiodomethane, to evaluate the different contributions to the surface energy. Furthermore, the wettability could be reversibly switched between hydrophobicity and hydrophilicity via alternation of UV exposure and dark storage. In addition, a more systematic study of the deposition condition’s effects, concerning, in particular, those determining samples morphology, seems to be necessary to exploit.
IV. SUMMARY
The spongy TiO layers deposited by a reactive sputtering method offer new possible applications in the field of engineering of nanomaterials, specifically for the photovoltaic field. This method is based on Grazing-Incidence Geometry coupled with Local OXidation (gig-lox) and is used for the growth of nanostructured TiO layers. It represents a new way to force sputtering methods, naturally addressed to compact layers, to arrange spongy materials via progressive bottom-up oxidation of building blocks. The technique has an industrial interest for its intrinsic upscalability and can be, in principle, extended to any reactive metallic source to produce porous oxides. The spongy gig-lox TiO is unique in its double structure: it consists of a regular array of rods, separated by mesopores (tens of nanometers) as a result of applying the grazing geometry. The rods, on their side, have an internal branched structure that gives rise to an interconnected network of nanopores (a few nanometers) related to the proper partial pressure of Ar used for sputtering. The overall bimodal porosity of the layer amounts to 50% of the volume for thicknesses tunable up to (at least) 1000 nm. The preliminary results show good practical perspectives in the photovoltaic field.
The increased wettability by applying a thermal budget on the TiO surface is an important result since in the preparation of perovskite solar cells by spin-coating, for example, high wettability can participate to achieve more uniform and higher quality samples. Future studies will be conducted in this field to systematically investigate the effects of the hydrophobic/hydrophilic ratio on our TiO-based solar cell performance. Furthermore, we have already planned the study of wettability variations on TiO samples grown on different substrates upon exposure to ultraviolet light over time.
ACKNOWLEDGMENTS
The authors wish to thank the FSE (Fondo Sociale Europeo) and the “Programma Operativo Nazionale” (PON) for Sicily 2014-2020. The authors also acknowledge the Mission Innovation Grant Agreement between the Italian Ministry of Ecological Transition and ENEA, Italian National Agency for New Technologies, Energy and Sustainable Economic Development (Ref. Agreement No. 21A03302 GU n.133 del 5-6-2021), CUP B82C21001820001.
AUTHOR DECLARATIONS
Conflict of Interest
The authors have no conflicts to disclose.
Author Contributions
C. Spampinato: Conceptualization (lead); Data curation (lead); Formal analysis (lead); Funding acquisition (lead); Investigation (lead); Methodology (lead); Project administration (lead); Resources (lead); Software (lead); Supervision (lead); Validation (lead); Visualization (lead); Writing – original draft (lead); Writing – review & editing (lead). S. Valastro: Writing – review & editing (equal). E. Smecca: Writing – review & editing (equal). V. Arena: Writing – review & editing (equal). G. Mannino: Writing – review & editing (equal). A. La Magna: Writing – review & editing (equal). C. Corsaro: Conceptualization (equal); Formal analysis (equal); Methodology (equal); Resources (equal); Software (equal); Supervision (equal); Writing – original draft (equal); Writing – review & editing (equal). F. Neri: Writing – review & editing (equal). E. Fazio: Conceptualization (equal); Formal analysis (equal); Methodology (equal); Resources (equal); Software (equal); Supervision (equal); Validation (equal); Visualization (equal); Writing – original draft (equal); Writing – review & editing (equal). A. Alberti: Conceptualization (equal); Formal analysis (equal); Methodology (equal); Resources (equal); Supervision (equal); Validation (equal); Visualization (equal); Writing – original draft (equal); Writing – review & editing (equal).
DATA AVAILABILITY
The data that support the findings of this study are available from the corresponding author upon reasonable request.