Solution-Processed Chalcogenide Photovoltaic Thin Films

*Marcos Antonio Santana Andrade Junior, Hugo Leandro Sousa dos Santos, Mileny dos Santos Araujo, Arthur Corrado Salomão and Lucia Helena Mascaro*

## **Abstract**

Chalcogenides-based thin film solar cells are great competitors to beat high efficiencies as silicone solar cells. The chalcogenides that have been commonly used as absorber materials are CIS, CIGS, and CZTS. They present some advantages of having a direct and tunable band gap, high absorption coefficient and respectable efficiency to cost ratio. Solution processable deposition approaches for the fabrication of solar cells attracts a great deal attention due to its lower capital cost of the manufacturing than the vacuum-based techniques. In this chapter, we detail the use of a low-cost method of deposition for the chalcogenide thin films by spin-coating and spray-coating, which is already widely employed in several fields of industries.

**Keywords:** chalcogenides, solar cells, spin-coating, spray

### **1. Introduction**

Currently, the photovoltaic market is based on silicon solar cells with conversion efficiency of 15–22%. However, chalcogenides-based thin films solar cells are great competitors of silicon technologies. Among of all photovoltaic thin films, chalcogenide solar cells present some advantages of having a direct and tunable band gap (1.0–2.5 eV), high absorption coefficient (>104 cm−1) and respectable efficiency to cost ratio [1]. Devices containing CuInGa(S,Se)2 (CIGS) recently recorded a 23% efficiency [2] and it becomes possible to PV modules based on that chalcogenide thin film to have already entered the market at similar or even lower costs than traditional silicon modules [3]. Despite this high efficiency level, the CIGS-based PV technology has not yet attained its full potential. If all loss mechanism were addressed at the same time, the theoretical maximum 30% could be technically feasible. The reported highest efficiency for CIGS thin film solar cell is based on vacuum process for coevaporating the elements and depositing the absorber layer. However, vacuum-based methods are quite expensive, which creates cost and technological barriers to produce low-cost photovoltaic modules [4]. Therefore, some approaches have been investigated for further improvements, as well as, to develop cheaper strategies for the absorber layer.

Solution-processed techniques have been extensively studied to deposit chalcogenide thin films applied to the second-generation solar cells. The characteristics of the precursor solutions are fundamental to perform the deposition and have an important role in the resulting film. The solution composition, concentration

of constituents, viscosity, and solvent will influence on the film adherence onto substrate, grain growth, and most importantly, on the solar cell efficiency.

The solution processable deposition of chalcogenides absorber layer, such as CIS (CuIn(S,Se)2), CZTS (Cu(Zn,Sn)(S,Se)2), and CIGS compounds for the fabrication of solar cells attracts a great deal attention due to its lower capital cost of the manufacturing than the vacuum-based approaches, high production rate, ability for roll-to-roll production, compositional uniformity over large areas, and high material utilization [5, 6]. Several solution processed methods are already commonly used for chalcogenides film deposition, such as, electrodeposition [4], spin coating [7], ink jet printing, electrophoretic deposition [8], etc.

Among the solvents used to prepare solution-processed high-performance solar cells reported in literature are water (H2O), dimethyl sulfoxide (C2H6OS), hydrazine (N2H4), methanol (CH3OH), dimethylformamide (C3H7NO) [9]. The solvent used in the precursor solution must be environmental-friendly, being able to dissolve the salts used as source of the chalcogenide cations and the compounds that are the source of the chalcogens, and not contribute to impurities in the film [10].

This chapter discusses the two most reported techniques used to prepare the world-record efficiencies of solution-processed chalcogenide solar cells: spin-coating and spray-coating. The fundamental characteristics of the solutions, regarding the physical chemical properties of the solvents, and the important characteristics of the deposition methods will be discussed.

#### **2. Absorber layers deposited by spin-coating**

The spin coating is a technique to deposit thin films on flat substrates. It has been very utilized on the fabrication of films in thickness range of micrometer to nanometer. This makes it attractive to prepare solar cells, on a manufactory point of view, mainly at laboratorial scale.

The process to produce thin films with this method can be resumed basically in three steps (**Figure 1** represents a schematic diagram of the steps), respectively; fluid dispense on the substrate, spreading of fluid (spin up followed by spin off) and evaporation. The solution dispensed spreads over the substrate surface by the centrifugal force and, at last, the evaporation forms the film.

The viscosity of the fluid dispensed (η), density of volatile liquid (ρA), evaporation rate (*m*), and angular speed of the spinning plate (ω) are the main factors that

**Figure 1.** *Steps of spin coating technique.*

affects the layer thickness, the following equation represents the variation of the thickness in function of this parameters [11]:

$$h = \left(1 - \frac{\rho\_A}{\rho\_{A0}}\right) \left(\frac{3\eta.m}{2\rho\_{A0}a^2}\right)^{1/3} \tag{1}$$

where thickness is *h*. These factors and possible contaminants (oxygen, humidity, solvent traces, etc.) can also influence on the roughness and the uniformity of the film [12].

Spin coating is a fast and low-cost process. The feasibility of the process is due to cheap material required, comparing to vacuum-based coatings, since a spinning plate is much cheaper than vacuum system [12]. In addition, this method allows the use of a wide range of particle diameters, which means that the method can be utilized for different applications. On the solar cells, smaller particles diameters mean more grain boundaries, and consequently, loss of charge carrier [13].

Otherwise, it is important to note that at the spin coating process, there is an expressive material wastage, since only 2–5% of the material stays on the layers, while 95–98% drops out of the layer [14]. Another disadvantage is that this technique does not form uniform layers over a large area and it does not have roll-to-roll capability [15]. Lastly, there is a difficulty at the production of multilayers devices with this method.

The spin coating has a good cost effective, mainly at laboratorial use, to produce thin and uniform films. Anyway, it is indispensable to consider the advantages and the disadvantages before using this method to conclude if this is viable.

Since the success of the deposition of Sn(S,Se)x films using hydrazine-based solution, this method has been applied to fabricate highly efficient devices. CIGS thin film was deposited by spin coating using a hydrazine-based solution containing the chloride salts of the cations and thiourea [16]. Hydrazine is a solvent that has been used to prepare molecular inks to deposit highly pure films for highly efficient solar cells. The current world record efficiency for a solution-processed solar cell is hold by a device containing CIGS deposited via hydrazine-based solution (12.6% for a CIGS solar cell) [17, 18]. Hydrazine is used to dissolve elemental metals and binary compounds, as SnSe to produce kesterite solar cells [19]. It has also been used to prepare precursor solutions containing Cu2S e In2Se3 to deposit CuIn(S,Se)2 thin films [20]. The advantage of using hydrazine is because it is a poor coordination solvent. During thermal treatment, hydrazine decomposes in small molecules (H2, NH3 and N2) which easily leaves the film without keeping residues. The absence of carbon or oxygen in hydrazine structure avoid organic impurities which compromise the solar cell efficiency [21]. However, hydrazine use is limited because it is highly toxic irritant, highly reactive and easily catch fire [12]. Because of that, hydrazine-free solutions have been studied to unexpected dangerous reactions.

Alcohols are a green, less toxic, and low-cost alternative for deposition solutions. Methanol-based solution containing the chloride salts of the cations and thiourea, as sulfur source, have been used to deposit CIGS films. The cation acetate salts are more recommended rather than the nitrate and chloride-based salts to form absorber layers with a minimum of residues.

Alcoholic solutions have also been used to prepare inks with butyldithiocarbamic acid (BDCA) as the sulfur source instead of thiourea to deposit CZTS solar cells with 6% conversion efficiency. Ethanol-based solutions with BCDA to dissolve binary oxides Cu2O, ZnO and SnO to deposit CZTS films [20]. Ammonium thioglycolate has also been used to dissolve metallic oxides as an environmentally friendly alternative [10].

The cations and thiourea complexes with dimethyl sulfoxide (DMSO) or dimethyl formamide (DMF) to prepare molecular inks to deposit chalcogenide absorber layers [22]. This combination acts stabilizing the desired oxidation state of the metals in the resulting film [10]. Thiourea binds to the cations and avoid their evaporation during annealing [23, 24]. DMF and DMSO also dissolves selenourea to prepare the selenide compounds [9].

Beyond the salt-based precursor solutions and molecular inks, nanocrystals inks have also been used to deposit chalcogenide absorber films. CZTS solar cells fabricated using a nanoparticle-based solution have presented a maximum of 10.2% conversion efficiency, which is close value to the efficiency for a device prepared using hydrazine. The synthesis of the nanocrystals typically consist of oleylamine solution of the cation salts (chlorides or acetylacetonate) [13, 25], and a hot injection of sulfur oleylamine solution. Dodecanethiol and hexanethiol are alternatives to oleylamine [20]. Dichlorobenzene has also been used to dissolve sulfur. Dichlorobenzene is not recommended to prepare Se-based chalcogenides because of the Se low solubility in this solvent [25]. After separation and purification, the nanocrystals are dissolved in an organic solvent such ethanol, hexane, chloroform, or toluene to prepare the nanocrystal inks.

The combination of molecular and nanocrystal inks is known as hybrid ink [20]. A hybrid solution prepared by dispersing Cu2SnS3 nanocrystals in a solution containing Zn and propylmercaptan to obtain a hybrid solution applied to deposit Cu2ZnSnS4 thin films. The hybrid solution-based CZTS solar cell presented a PCE of 5% and Voc of 440 mV [20]. Solvents may leave residues trapped in the film. The time and temperature for annealing and sulfur/selenization are not only essential for crystal growth but must also remove all the possible residues left by solvent and other organic [5, 26].

Although the CIGS devices prepared by vacuum-based techniques present values of efficiencies of ~23%, the efficiency of a spin-coating deposited CuIn(S,Se)2 solar cell is similar to the efficiency of a Cu(Ga,In)(S,Se)2 device fabricated in the same condition. In addition, the efficiency of a CZTS solar cell is the same as for a device fabricated by vacuum techniques. This evidences that spin-coating is an inexpensive alternative to fabricate highly efficient devices. The efforts for developing inks using environmentally friendly solvents, metals and chalcogen precursors, decreasing the residues in the films are fundamental to approximate the photovoltaic characteristics of a solution-processed device to the characteristics of the expensive vacuum-based solar cells (**Table 1**).


#### **Table 1.**

*Photovoltaic parameters of highly efficient chalcogenide solar cells fabricated by spin-coating.*
