Chemical Solution Deposition

Chemical solution deposition (CSD) is one of the most common processes for the fabrication of thin films, specifically oxide films . It consists of the deposition of a solution onto a surface during which chemical reactions occur in order to synthesize the desired material. The final crystalline material is achieved after a heat treatment. CSD was originally developed in the mid 1980s . Fukushima and Budd were two of the early engineers who studied the CSD process and made important findings . They showed that it was possible to achieve bulk material properties in ferroelectric thin films made by this process ^[2,3]. CSD consists of three main processing steps .

A. Synthesis of the Precursor Solution: Combining the essential chemical ingredients required for the formation of the desired phase and appropriate to the chemical process chosen; all the necessary cations that are required to synthesize the material must be present in a homogeneous solution ^[3]

B. The Coating or Deposition Process: Creating the homogeneous precursor layers on the desired substrate

C. Heat Treatments: Converting the homogeneous, amorphous as deposited layers into the final crystallized phase

There are many advantages CSD has over other thin film deposition techniques, such as chemical vapor deposition (CVD), physical vapor deposition (PVD), sputtering, or molecular beam epitaxy (MBE). In chemical solution deposition, control of the exact composition on a molecular level (local stoichiometry) is possible ^[1,3]. It is easy to incorporate multiple elements into the solution, which is a large advantage when processing complex oxide films . CSD is one of the simplest and least expensive thin film fabrication processes because typically no vacuum system or elaborate deposition chamber is required. Low temperature synthesis of the thin films is possible as well as being able to coat a large deposition area ^[1,5]. The vaporization or ionization of the solutions is not required for deposition. It produces films of high purity and good homogeneity ^[1]. CSD is a process which can be modified quickly in order to synthesize a new material system that may not have been studied in a thin film form ^[3].

CSD also has some disadvantages compared to conventional thin film processing. It is difficult to deposit layers less than 30nm thick. Usually only polycrystalline films are grown with this method. It is more difficult to grow oriented or epitaxial films ^[3]. CSD is GeneRally not found in manufacturing.

A. Synthesis of the Precursor Solution

There are multiple routes that can be taken to prepare the chemical solution. CSD can be separated into different categories. For the preparation of ferroelectric materials the three categories include metal-organic decomposition (MOD), sol-gel, and chelation ^[3]. According to some scientists chelation is categorized as a form of the sol-gel method because it is quite similar ^[4]. Therefore the sol-gel process can be subcategorized into different routes depending on the exact chemistry involved. Modified versions include the chelate process, the nitrate-citrate route, and the Pechini method ^[4]. Each CSD route can be characterized by the chemical reaction and the type of precursors used to make the solution.

Most CSD solutions are created using metal-organic compounds as the main source of cations. Before discussing the different CSD routes, a quick explanation of the different metal-organic precursors used will be given. The main solution precursors can be divided in to three classes ^[3].

a. Alkoxides – metal salts of alcohols; M(OR)x, where M = metal, OR = alkyl group The structure of an alkoxide is dimeric, meaning each molecule has two metal centers. These compounds are very sensitive to water. Their M-O bond likes to transform into M-OH. Alkoxides are used to form oligomers, which are short-chained polymers. The very sensitive alkoxide can be stabilized to not be as sensitive to water. When more polar groups are present around the alkyl, more bonds can form to the metal cation. Organic chemicals used to bond to the alkyl to form R-O-R bonds are 2MOE = 2-methoxyethanol, amine groups (-NH₂), keto groups (-C=O), and alcohol groups (-C-OH). All of these are polar groups meaning there is a dipole present. When multiple polar groups bond to the metal, the groups are referred to as a chelating agent. When a chelating agent is used, the alkoxide becomes even more stable and less sensitive to water ^[3].

b. Carboxylates – salts of carboxylic acids; R-COOH, where R=alkyl Caboxylates, in general, are very stable in water and oxygen and are generally used in the MOD process. Examples of caboxylates are acetate and propionate. Their carboxylic acid counterparts are acetic acid (R=CH₃) and propionic acid (R=C₂H₅). It is easiest to dissolve these salts in their paired acid. For extra long-chained alkyl groups (C₇H₁₅), xylene can be used as the solvent ^[3].

c. Beta-diketonates – organic molecules where a methylene group (-CH₂) divides two keto groups

Beta-diketonates are molecules in which an exchange of a hydrogen atom occurs in order to transform a keto into an alcohol. These tend to dissolve easily in alcohols, ketones, and ethers. In general, beta-diketonates have the potential to produce denser films thans alkoxides and carboxylates. Films made with these precursors require higher pyrolysis temperatures and tend to remain amorphous up to higher temperatures ^[3].

Metal-Organic Decomposition

Metal-organic decomposition uses a solution of metal organics that are insensitive to hydrolysis ^[4]. Common metal organics are carboxylate compounds such as acetates and ethyl hexanoates or beta-dikeytonates such as pentanediaonates ^[4]. These are dissolved in nonpolar solvents ^[5], for example, xylene ^[2]. The metal organic compounds do not interact with the solvent and there is no oligomerization behavior as in sol-gel or chelate ^[2]. The solution is very stable and can be seen as a simple mixture of the starting ingredients ^[5]. The precursors remain representative of the starting molecules ^[2]. The precursor compounds are physically condensed onto the substrate through rapid solvent evaporation . The liquid film is then rapidly heat-treated immediately after deposition to force phase formation ^[1]. The reactions that occur to force phase formation occur by thermal decomposition during the pyrolysis or the burnout stage. The organic material is usually removed in the form of CO₂ or water ^[5]. Extreme exothermic reactions can occur during this stage that can cause biaxial tensile stresses in the film ^[5]. These stresses can cause cracking of the film ^[5].

Sol-Gel

Sol-gel is a chemical solution process used to make ceramic and glass materials in the form of thin films, fibers (extruded shapes), or powders ^[1]. In essence a liquid film is transformed into a solid through hydrolysis and polycondensation reactions ^[1]. A sol is a colloidal (the dispersed phase is so small that gravitational forces do not exist; only Van der Waals forces and surface charges are present) or molecular suspension of solid particles of ions in a solvent . A gel is a semi-rigid mass that forms when the solvent from the sol begins to evaporate and the particles or ions left behind begin to join together in a continuous network ^[7].

The sol typically consists of metal alkoxides (tetraethylorthosilicate=TEOS, titanium isopropoxide, aluminum butoxide) or metal salts (acetates, nitrates) ^[4,5,8]. The metal alkoxides are usually in the form of covalent liquids . These are dissolved in an alcohol based solvent to form a stable sol ^[5]. The alcohol promotes miscibility of the alkoxide and water ^[8]. Water is a key ingredient to cause the hydrolysis reaction ^[8]. Sol-gel is similar to MOD, except that the chemistry is designed to produce metal-oxide-metal chains using hydrolysis and polycondensation reactions ^[5]. The formation of the chains creates a 3D continuous gel network which is the amorphous oxide layer ^[5]. Thermal treatments are performed to turn the amorphous layer into the desired crystalline oxide phase.

The series of reactions that occur during the sol-gel process include:

i. Hydrolysis: M(OR)_x + H₂O → M(OR)_x-1(OH) + ROH; where M(OR)_x = alkoxide compound, M = metal, OR = alkyl group

ii. Condensation: (-OH elimination): 2M(OR)_x-1(OH) → M₂O(OR)_2x-3(OH) + ROH

iii. Condensation: (H₂O elimination): 2M(OR)_x-1(OH) → M₂O(OR)_2x-2(OH) + H₂O

Chelate Process

Like sol-gel, the chelate process (Chelation) utilizes a liquid solution and similar deposition techniques. This solution however is typically made from alkoxide compounds, like sol-gel, but consists of modifying ligands such as acetic acid, acetylacetone, or amines to cause the chemical reaction to occur ^[2,3]. The solution preparation time is much less than sol-gel; usually is requires less than one hour compared to sol-gel’s one to two day prep time and is also simpler ^[2,3]. No distillation or refluxing is required ^[2]. Like sol-gel, oligomerization occurs during the solution synthesis ^[3]. The solutions tend to more stable in air than sol-gel solutions because the chelate agents decrease the hydrolysis sensitivity of the alkoxide ^[3].

M(OR)_n + xCH₃COOH → M(OR)_n-x(OOCCH₃)_x + xROH

Nitrate-Citrate Process

This is a modified sol-gel process which is also called the citrate gel process ^[8]. The precursor ingredients include metal nitrates which are dissolved into citric acid and water ^[8]. This process is used to make YBa₂Cu₃O₇, a superconductor material ^[8].

Pechini Method

The Pechini method is a modified sol-gel process which uses an aqueous solution of polymeric precursors dissolved in an alpha-hydroycarboxylic acid (citric acid) and water . The chemical reactions which occur during the synthesis include chelation or the formation of ring compounds around the metal cations ^[9].

B. The Coating or Deposition Process

Deposition or coating of the chemical solutions onto their desired substrates can occur many different ways. The technique chosen usually is dependent on what type of substrate or object needs to be coated. Different techniques include spin coating, dip coating, spraying, roll coating, slot-die coating, and ink-jet printing ^[4]. For all these techniques the thickness of the precursor layer is dependent upon the solution's properties, such as molarity and viscosity ^[4]. Thickness is also dependent on factors from the deposition technique including, deposition speed and coating time ^[4]. Properties of the substrate also effect thickness. Surface tension and wetting play large rolls in coating ability and thickness.

Spin coating

a. Deposition – The liquid solution is dispersed on surface of substrate. Spin-coating is only capable of coating one side of a substrate unlike dip coating which can coat an entire non-flat object.^[1]

b. Spin-up – The liquid flows radially outward due to the centrifugal force resulting from the spinning.

c. Spin-off – The excess liquid flows off the substrate in droplets as the film becomes uniform.

d. Evaporate – The solvent leaves the film as it spins while the film’s viscosity and density increase. The film thins.

e. Pyrolysis – A heat treatment, typically under 500C, performed to remove the organics from the film in order to ensure densification during annealing step.

f. RTA – (Rapid Thermal Annealing) Crystallization and densification of the film. The film is soaked at high temperature for a brief amount of time (one minute or two).

Dip-coating

The substrate or object to be coated is lowered into the solution and withdrawn at a controlled speed. There is no special equipment required, making it a very simple and inexpensive method. It is capable of coating the inside and outside of the object.^[1]

C. Heat Treatments

The thermal treatments required to removed organics and crystallize the film are different for different solution routes and different materials. The heat treatments are required to transform the as deposited layer into its desired crystallize phase. There are two main stages or at least two separate heat treatments ^[4].

i. Pyrolysis - This stage is required in order to remove the organic matter that remains in the amorphous film. Typical temperatures for pyrolysis range from 300-600C. During this stage a large volume reduction occurs. Stress and microcracks can easily form. It is possible for an intermediate state to exist between the pyrolysis and the annealing stage. This state is an amorphous, nanocrystalline, metastable, porous film ^[4].

ii. Annealing - This stage causes the crystallization of the amorphous intermediate film state into its desired phase. The temperatures for this stage usually range between 600-1200C ^[4]. The exact temperature for a specific material can be determined by examining the phase diagram. During the anneal many of the stresses that formed during pyrolysis are removed.

Spin Spray Plating Technique

(unconventional CSD, may not exactly be considered)

The spin spray plating technique is used to synthesize thin films of the spinel crystal structure at atmospheric pressure, no vacuum system required, and at temperatures under 100C with no additional heat treatments required ]. A research group at the Tokyo Institute of Technology developed the technique and the equipment needed to deposit these films ^[11].

The plating process is based on the treatment of wastewater. Wastewater can be treated by a ferrite process in which the heavy metal ions in the water would be incorporated into a spinel lattice by a wet process ^[11]. The wet process would create powders that could then be separated from the water by magnetic separation ^[11].

Two aqueous solutions are deposited simultaneously on a substrate mounted on a heated, spinning table in a nitrogen-rich environment. The two solutions are:

a. Reaction Solution – Consists of the metal chlorides that are desired for the spinel structure

b. Oxidizing Solution – Consists of ammonium acetate and sodium nitrite

The substrate begins with –OH groups on its surface. The reaction solution containing the metal chlorides is sprayed onto the heated substrate. If the desired spinel was Fe₃O₄, the Fe²⁺ ions bond to the O^2- while the H⁺ is removed. The oxidizing solution is sprayed on and causes the Fe²⁺ to oxidize and become Fe³⁺ while releasing an electron. The reaction continues and forms layer by layer a ferrite spinel structure (Fe₃O₄). If another cation was present in the reaction solution—for example Co²⁺ → (CoFe₂O₄).