Motivation

In this paper, we will discuss the optical and electric characterizations of GZO thin film deposited by APPCVD, because most researches with APP systems are investigated into mechanical mechanism such as a corrosion protection layer, a hard coating material, a passivation layer for the case of polymers and food packing, etc. In fact, GZO has high transparency in visible (＞80%) and low resistivity (<10^-3Ωcm) by other deposition ways. With the advantages of APPCVD for instance, low temperature, atmospheric environment, high deposition rate, low cost, and adaption of large substrates. Therefore, we will try to deposit GZO by APPCVD process to get good TCO films.

Fig 1-1 Cost of Indium in recent year [4]

Table 1-1 Application of the TCO thin films [4]

Chapter 2

Literature Review

2.1 Properties of ZnO

Zinc oxide is an inorganic compound with the formula ZnO. It usually appears as a white powder, nearly insoluble in water. The powder is widely used as an additive into numerous materials and products including plastics, ceramics, glass, cement, rubber (e.g., car tires), lubricants, paints, ointments, adhesives, sealants, pigments, foods (source of Zn nutrient), batteries, ferrites, fire retardants, first aid tapes, etc. ZnO is present in the Earth's crust as the mineral zincite; however, most ZnO used commercially is produced synthetically.

In material science, ZnO is often called a II-VI semiconductor because zinc and oxygen belong to the 2nd and 6th groups of the periodic table, respectively. This semiconductor has several favorable properties:

good transparency, high electron mobility, wide bandgap, strong room temperature luminescence, etc. Those properties are already used in emerging applications for transparent electrodes in liquid crystal displays and in energy-saving or heat-protecting windows, and electronic applications of ZnO as thin-film transistors.

Zinc oxide crystallizes in three forms: hexagonal wurtzite, cubic zincblende, and the rarely observed cubic rocksalt. The wurtzite structure is most stable at ambient conditions and thus most common. The zincblende form can be stabilized by growing ZnO on substrates with

cubic lattice structure. In both cases, the zinc and oxide centers are tetrahedral. The rocksalt (NaCl-type) structure is only observed at relatively high pressures about 10 GPa.

Hexagonal and zincblende polymorphs have no inversion symmetry (reflection of a crystal relatively any given point does not transform it into itself). This and other lattice symmetry properties result in piezoelectricity of the hexagonal and zincblende ZnO, and in pyroelectricity of hexagonal ZnO.

The hexagonal structure has a point group 6 mm (Hermann-Mauguin notation) or C6v (Schoenflies notation), and the space group is P63mc or C6v4. The lattice constants are a = 3.25 Å and c = 5.2 Å; their ratio c/a ~ 1.60 is close to the ideal value for hexagonal cell c/a = 1.633. As in most group II-VI materials, the bonding in ZnO is largely ionic, which explains its strong piezoelectricity. Due to the polar Zn-O bonds, zinc and oxygen planes bear electric charge (positive and negative, respectively).

Therefore, to maintain electrical neutrality, those planes reconstruct at atomic level in most relative materials, but not in ZnO - its surfaces are atomically flat, stable and exhibit no reconstruction. This anomaly of ZnO is not fully explained yet.

(a) (b)

Figure 2-1 Two views of the crystal structure of zinc oxide: (a) Perspective view perpendicular to the c-axis. (b) View along the c-axis on the zinc terminated (0001) plane [5].

ZnO has a relatively large direct band gap of ~3.3 eV at room temperature; therefore, pure ZnO is colorless and transparent. Advantages associated with a large band gap include higher breakdown voltages, ability to sustain large electric fields, lower electronic noise, and high-temperature and high-power operation. The band gap of ZnO can further be tuned from ~3–4 eV by its alloying with magnesium oxide or cadmium oxide.

Most ZnO has n-type character, even in the absence of intentional doping. Nonstoichiometry is typically the origin of n-type character, but the subject remains controversial. An alternative explanation has been proposed, based on theoretical calculations, that unintentional substitutional hydrogen impurities are responsible. Controllable n-type doping is easily achieved by substituting Zn with group-III elements such as Al, Ga, In or by substituting oxygen with group-VII elements chlorine or iodine.

Reliable p-type doping of ZnO remains difficult. This problem originates from low solubility of p-type dopants and their compensation by abundant n-type impurities. This problem is observed with GaN and ZnSe. Measurement of p-type in "intrinsically" n-type material is complicated by the inhomogeneity of samples.

A current limitation to p-doping does not limit electronic and optoelectronic applications of ZnO, which usually require junctions of n-type and p-type material. Known p-type dopants include group-I elements Li, Na, K; group-V elements N, P and As; as well as copper and silver. However, many of these form deep acceptors and do not produce significant p-type conduction at room temperature.

2.2 CVD Mechanism 2.2.1 Theory of CVD

Chemical vapor deposition (CVD) is defined as the formation of a nonvolatile solid film on a substrate by the reaction of vapor-phase chemicals (reactants) that contain the required constituents. It is most often used for semiconductor processing. It is a material synthesis process whereby the constituent of the vapor phases react chemically near or on a substrate surface to form a solid product.

Several steps must occur in every CVD Reactions: [6]

z Gas or vapor phase precursors are introduced into the reactor.

z Precursors diffuse across the boundary layer and reach the substrate surface.

z Precursors adsorb on the substrate surface.

z Adsorbed precursors migrate on the substrate surface.

z Chemical reaction begins on the substrate surface.

z Solid byproducts form nuclei on the substrate surface.

z Nuclei grow into islands.

z Island merge into the continuous thin film.

z Other gaseous byproducts desorb from the substrate surface.

z Gaseous byproducts diffuse across the boundary layer.

z Gaseous byproducts flow out of the reactor.

The sequence of reaction steps in a CVD process is illustrated in Fig 2-2. In practice, the chemical reactions of the reactant gases leading to the formation of a solid material may take place not only on (or very close to) the wafer surface (heterogeneous reaction) but also in the gas phase (homogeneous reaction). Heterogeneous reactions are much more desirable, because such reactions occur selectively only on the heated

surfaces and produce good-quality films. Homogeneous reactions on the other hand, are undesirable, because they form gas-phase clusters of the depositing material, resulting in poorly adhering, low-density films with defects. In addition, such reaction also consume the reactants and cause a decrease in deposition rates. Thus, one important issue of a chemical reaction for CVD application is the degree to which heterogeneous reactions are favored over gas-phase reactions.

Figure 2-2 Schematic of CVD process sequence [6]

Reactants Byproducts

Showerhead

Pedestal Wafer

Precursors

Forced convection region

Boundary layer

Figure 2-3 Schematic of CVD reaction steps [6]

Island grow (Cross section)

Island merge Film Island grow

Nucleation Reaction

Migration

2.2.2 Chemical Reaction Rate

The Arrhenius equation gives the quantitative basis of the relationship between the activation energy and the rate at which a reaction proceeds. From the Arrhenius equation, the activation energy can be expressed as the following:

K= A exp (-Ea / RT) (Eq.

2-1)

where A is the frequency factor for the reaction, R is the universal gas constant, T is the temperature (in kelvins), k is the reaction rate coefficient, and Ea is the activation energy, as illustrated in Fig 2-2. While this equation suggests that the activation energy is dependent on temperature, in regimes in which the Arrhenius equation is valid this is cancelled by the temperature dependence of k. Thus Ea can be evaluated from the reaction rate coefficient at any temperature (within the validity of the Arrhenius equation). The lower the activation energy E_a’ the easier the chemical reaction. External energy sources such as heat, RF power, or UV radiation are needed for the chemical precursors to overcome the activation energy barrier and achieve the chemical reaction.

Because the chemical reaction rate is exponentially related to the temperature, it is very sensitive to change of temperature. Changing temperature can dramatically change the chemical reaction rate. For a CVD process, the deposition rate (D.R.) is related to chemical reaction rate (C.R.), the precursor diffusion rate in the boundary layer (D), and the precursor adsorption rate on the substrate (A.R.)

From Fig 2-4 one can see that the deposition rate has three regimes

when the temperature changes. At lower temperatures, the chemical reaction rate is low, and the deposition rate is very sensitive to the temperature. This is called the surface-reaction-limited regime. At higher temperatures, the deposition is much less sensitive to the temperature.

This is the mass-transport-limited regime. When the temperature further increases, the deposition rate sharply decreases due to the gas phase nucleation. This is a very undesirable deposition region, since the chemical precursor react in midair, generating huge amount of particles and contaminating the wafer and the reactor. The gas-phase-nucleation or homogenous-nucleation regime must be avoided for all CVD process in IC thin-film depositions.

Fig 2-4 Chemical activation energy. [6]

2.3 Plasma

2.3.1 Definition of Plasma

The plasma state is the fourth state of matter. Plasma is an ionized gas with equal numbers of positive and negative charges. A more precise of plasma is a quasi-neutral gas of charged and neutral that exhibits collective behavior.

In physics and chemistry, plasma is a gas in which a certain portion of the particles are ionized. The presence of a non-negligible number of charge carriers makes the plasma electrically conductive so that it responds strongly to electromagnetic fields. Plasma therefore has properties quite unlike those of solids, liquids, or gases and is considered to be a distinct state of matter. Like gas, plasma does not have a definite shape or a definite volume unless enclosed in a container; unlike gas, in the influence of a magnetic field, it may form structures such as filaments, beams and double layers. Some common plasmas are flame, lightning, and the Sun.

Fig 2-5 Density and Energy for Various Species in a Low-Pressure Capacitive RF Discharge (RIE).

2.3.2 Glow Discharge

The simplest type of glow discharge is a direct-current glow discharge. In its simplest form, it consists of two electrodes in a cell held at low pressure (0.1–10 torr; about 1/10000th to 1/100th of atmospheric pressure). The cell is typically filled with argon, but other gases can also be used. An electric potential of several hundred volts is applied between the two electrodes. A small fraction of the population of atoms within the cell is initially ionized through random processes (thermal collisions between atoms or with alpha particles, for example). The ions (which are positively charged) are driven towards the cathode by the electric potential, and the electrons are driven towards the anode by the same potential. The initial population of ions and electrons collides with other atoms, ionizing them. As long as the potential is maintained, a population of ions and electrons remains.

Some of the ions' kinetic energy is transferred to the cathode. This happens partially through the ions striking the cathode directly. The primary mechanism, however, is less direct. Ions strike the more numerous neutral gas atoms, transferring a portion of their energy to them.

These neutral atoms then strike the cathode. Whichever species (ions or atoms) strike the cathode, collisions within the cathode redistribute this energy until a portion of the cathode is ejected, typically in the form of free atoms. This process is known as sputtering. Once free of the cathode, atoms move into the bulk of the glow discharge through drift and due to the energy they gained from sputtering. The atoms can then be collisionally excited. These collisions may be with ions, electrons, or other atoms that have been previously excited by collisions with ions, electrons, or atoms. Once excited, atoms will lose their energy fairly quickly. Of the various ways that this energy can be lost, the most

important is radiatively, meaning that a photon is released to carry the energy away. In optical atomic spectroscopy, the wavelength of this photon can be used to determine the identity of the atom (that is, which chemical element it is) and the number of photons is directly proportional to the concentration of that element in the sample. Some collisions (those of high enough energy) will cause ionization. In atomic mass spectrometry, these ions are detected. Their mass identifies the type of atoms and their quantity reveals the amount of that element in the sample.

Figure 2-6 shows the main regions that may be present in a glow discharge. Regions described as "glows" emit significant light; regions labeled as "dark spaces" do not. As the discharge becomes more extended (i.e., stretched horizontally in the geometry of the figure), the positive column may become striated. That is, alternating dark and bright regions may form. Relatedly, compressing the discharge horizontally will result in fewer regions. The positive column will be compressed while the negative glow will remain the same size, and, with small enough gaps, the positive column will disappear altogether. In an analytical glow discharge, the discharge is primarily a negative glow with dark region above and below it.

Below the ionization voltage or breakdown voltage there is no glow, but as the voltage increases to the ionization point the Townsend discharge happens just as glow discharge becomes visible; this is the start of the normal glow range. As the voltage is increased above the normal glow range, abnormal glow begins. If the voltage is increased to the point the cathode glow covers the entire cathode arc discharge begins.

Fig 2-6 Electric glow discharge tube.

2.3.3 Plasma Reactions [6]

In general, the plasma system has four important parts of reactions:

ionization, excitation-relaxation, dissociation and Recombination. All of them are inelastic collision. All the important reactions are listed as followings.

1. Ionization

When an electron collides with an atom or a molecule, it can transfer part of its energy to the orbital electron confined by the nucleus of an atom or molecule. If the orbital electron gains enough energy to break free from the constraint of the nucleus, it becomes a free electron. This process is called electron-impact ionization.

Ionization collision can be express as:

e^- + A ↔ A⁺ + 2e^- (Eq. 2-2)

Here e^- represents an electron, A represents a neutral atom or molecule, and A+ represents a positive ion. Ionization is very important

because it generates and sustains the plasma.

2. Excitation-Relaxation

Sometimes the orbital electron does not get enough energy from the impact electron to break free from the constraint of the nucleus.

However, if the collision transfers enough energy to the orbital electron to jump to a higher energy level of the orbit, it will do so.

This process is called excitation. It can be expressed as:

e^- + A ↔ A* + e^- (Eq. 2-3)

Here A* is the excited state of A, which indicates it has an electron in the higher energy level orbit very long, and will fall back to the orbit with the lowest possible energy level, or ground state.

This process is called relaxation. The excited atom or molecule quickly relaxes back to its ground state and releases the extra energy it gained from the electron impact in the form of a photon, which is the light emission.

A* ↔ A + hν (photons) (Eq. 2-4)

Here hν is the energy of the photon, h is Planck’s constant, and ν is the frequency of the light emission, which determines the color of light emitted from the plasma. Different atoms or molecules have different orbital structures and energy level; therefore, the light emission frequencies differ. That is why different gases glow in various colors in the plasma. The glow of oxygen is grayish-blue, nitrogen is pink, neon light is red, fluorine glow is orange-red, etc.

3. Dissociation

When an electron collides with a molecule, it can break the chemical bond and generate free radicals if the energy transferred by

the impact to the molecule is higher than the molecule bonding energy. The dissociation collision can be express as:

e ^- + AB → e^- + A + B (Eq. 2-5) Here AB is a molecule, and both A and B are the free radicals generated by dissociation collision. Free radicals are molecular fragments with at least one unpaired electron, which makes them unstable. Free radicals are chemically very reactive since they have a strong tendency to grab an electron from another atom or molecule to form a stable molecule. Free radicals can enhance chemical reaction in both etch and CVD processes.

4. Recombination

Recombination is the process during which the electron and ions are combined into a neutral atom or molecule. Recombination of positive and negative ions is also possible in electronegative plasmas.

Among the aforementioned processes the three-body collision is far more important. Actually, when plasma becomes stable, the recombination rate is equal to the ionization rate. Minimizing recombination rate useful for promoting plasma density. All the reaction of recombination are listed as followings:

Direct recombination:

e ^- + A⁺→ A (Eq. 2-6) A positive ion collides with an electron or negative ions

Recombination by three-body collision:

e ^- + A⁺ +B → A + B (Eq. 2-7) (B can be a gas atom, electrode, or the chamber wall)

Radiative recombination:

e ^- + A⁺ → A + hν (photons) (Eq. 2-8)

2.4 Atmospheric-pressure plasma system 2.4.1 Corona Discharge

A corona is a process by which a current, perhaps sustained, develops from an electrode with a high potential in a neutral fluid, usually air, by ionizing that fluid so as to create plasma around the electrode. The ions generated eventually pass charge to nearby areas of lower potential, or recombine to form neutral gas molecules.

When the potential gradient is large enough at a point in the fluid, the fluid at that point ionizes and it becomes conductive. If a charged object has a sharp point, the air around that point will be at a much higher gradient than elsewhere. Air near the electrode can become ionized (partially conductive), while regions more distant do not. When the air near the point becomes conductive, it has the effect of increasing the apparent size of the conductor. Since the new conductive region is less sharp, the ionization may not extend past this local region. Outside of this region of ionization and conductivity, the charged particles slowly find their way to an oppositely charged object and are neutralized.

If the geometry and gradient are such that the ionized region continues to grow instead of stopping at a certain radius, a completely conductive path may be formed, resulting in a momentary spark, or a continuous arc.

Corona discharge usually involves two asymmetric electrodes; one highly curved (such as the tip of a needle, or a small diameter wire) and one of low curvature (such as a plate, or the ground). The high curvature ensures a high potential gradient around one electrode, for the generation of plasma.

Coronas may be positive or negative. This is determined by the polarity of the voltage on the highly-curved electrode. If the curved electrode is positive with respect to the flat electrode we say we have a positive corona, if negative we say we have a negative corona. (See below for more details.) The physics of positive and negative coronas are strikingly different. This asymmetry is a result of the great difference in mass between electrons and positively charged ions, with only the electron having the ability to undergo a significant degree of ionising inelastic collision at common temperatures and pressures.

An important reason for considering coronas is the production of ozone around conductors undergoing corona processes. A negative corona generates much more ozone than the corresponding positive corona.

Figure 2-3 (a) shows a schematic of a point-to-plane corona.

Figure 2-3 (a) shows a schematic of a point-to-plane corona.