In the above mentioned, the three step co-evaporation technique has reached a confirmed photovoltaic energy conversion efficiency of about 20% [28]. This technique is limited to small area cells as evaporation from elemental sources was not suitable for large area coatings as composition uniformity cannot be maintained. And then co-evaporation technique has been the large amount of Se material used. These reason results in this technique had high cost compare to another technique. In this essay, we use the selenization of stacks of metals and compound
12
precursors for fabricated the large-area CIGS absorber layer. On the other hand, sputtering is widely known to be a technique easily scalable processes which are required for low-cost, large-area thin films solar cell production.
The selenization of sputtered Cu-In-Ga precursors using H2Se vapor has been proven to be a suitable method for the industrial production of Cu(In,Ga)2Se (CIGS)-based devices with the low cost. This technique has good control of the film growth, and result in CIGS based device of 16.2% [30]. One critical point is the use of high toxic H2Se, which is problematic in terms of economic and ecological aspects. So some papers begin to research the elemental Se vapor as an alternative over the past few years [31-35]. Therefore, we choice the Se vapor as the source of deposition Se layer. And then we will use the Atmosphere Pressure Plasma Enhance Chemical Vapor Deposition (APPECVD) to assist in depositing cap Se layer. The APPECVD system uses the Dielectric-Barrier Discharge (DBD) linear plasma source show in Figure 1-12. Therefore, this technique has an ability of deposition large area and develops fabrication of roll to roll in the future. We used high voltage to dissociate selenium molecule, and molecule group become smaller than before. This system not only get the smaller molecule group but also produce the Se radical when the Se vapor through the atmosphere pressure plasma. We can deposit the dense Se film on the CIG precursor layer due to the smaller molecule group in atmosphere pressure plasma system. And then the Se radical will assist the deposition selenium films to improve dense and roughness. Additionally, the plasma could
13
successfully dissociate the selenium vapor, and cause to amorphous selenium films on the glass and Mo/glass. We expected the degree of crystallinechalcopyrite structure of CIGS films was improved by radical Se and then improved conversion efficiency of device. Simultaneously, if the substrate temperature was high, we could expect the selenium films to crystalize oriented (101).
The selenization system has two categories include close-space furnace and Rapid Thermal Processing (RTP). In this paper, we will use the RTP to selenization process and Ar gas flux in to the chamber. This technique significantly reduces the thermal budget of the sample compared to conventional furnace annealing. These advantages of low thermal budget processing have to be noted such as minimization of inter-diffusion and impurity diffusion from the substrate as well as better control of the process kinetics.
14
Figure 1-1 Major PV country markets (GigaWatts).
Source: NPD Solarbuzz 2012 Marketbuzz.
Figure 1-2 Energy Band Diagram of a P-N Junction under Illumination.
15
Figure 1-3 I-V Characteristics under Dark and Illumination.
Figure 1-4 Inverted I-V Curve Showing Maximum Power Rectangle.
16
Figure 1-5 Shockley-Queisser Limits.
Figure 1-6 Absorption Coefficient versus Wavelength of Incident Light.[36]
17
Figure 1-7 The Unit Cell of the Chalocopyrite Lattice Structure.[10]
Figure 1-8 Ternary Phase Diagram of the Cu–In–Se System.[36]
18
Figure 1-9 Pseudobinary Cut Cu2Se-In2Se3 of Ternary Phase Diagram.[36]
Figure 1-10 Vacuum Processes for Cu(In,Ga)Se2.[29]
19
Figure 1-11 Non-Vacuum Processes for Cu(In,Ga)Se2.[29]
Figure 1-12 Dielectric-Barrier Discharge Linear Plasma Source.
High Voltage electrode
Dielectric Barrier Discharge
Ground Electrode High
Voltage AC Generator
20 Thermal conductivity at 273 K 0.086
Dielectric constant Low frequency 13.6 2.4 High frequency 8.1 1.4
Table 1-2 The Most Important Intrinsic Defects for Device-Quality CuInSe2.
Native Point Defect Electrical Activity Cui, SeCu, InSe Single dornor
21
Chapter 2
Literature Reviews
2.1 The Material Properties of Selenium
Selenium is a semiconductor with the property of conducting electricity better in the light than in the dark, and is used in photocells. So the absorber layer of CIS/CIGS solar cell will use selenium element to combine quaternary compound. The basic properties of selenium are show in the Table 2-1. The amorphous selenium has two allotropic forms including black and red. The black amorphous selenium is citreous and is formed by rapid cooling of liquid selenium. The red amorphous selenium is colloidal and is formed in reduction reactions. The crystalline selenium has several allotropic forms due to the difference of crystalline structure.
The stable form at room temperature is gray and the structure is hexagonal. This crystalline selenium is the densest and similar metallic in appearance. Crystalline red selenium exists in two monoclinic forms obtained by evaporation of carbon disulfide extracts of amorphous red selenium. The α-monoclinic form has a unit cell formed of four puckered Se8 ring molecules. The β-monoclinic form is also made up of puckered Se8 rings. The amorphous forms and both monoclinic crystalline forms transform to hexagonal forms gray selenium by heat. The saturated vapor pressure is given
(2-1)
, where T is absolute temperature (K) and P is saturated vapor pressure (kPa) [37]. The primary species are Sen(2<n<8) in the selenium vapor at
22
below 900ºC. Above mentioned data is arranged show in Table 2-2.
2.2 The Effect of Plasma Selenium
In 1994, S. T. Laskshimikumar and A. C. Rastogi use plasma to assisted two stage selenization process for the preparation of selenide semiconductor thin films using elemental selenium vapor. In this paper, the elemental selenium was used to produce the Selenium vapor. The reactivity of the selenium itself was enhanced by creating ionized selenium species in the vicinity of the metal precursor using RF radiation for the preparation of selenide semiconductor films. The Se2 and Se of the Se vapor were created by the interaction with the RF radiation, which was the highly reactive element. Using this process, the semiconducting selenide films became highly crystalline and oriented at modest reaction temperature (250℃) even with short reaction time (2-5 min). In contrast, the crystalline without plasma treatment was smaller than with plasma treatment at higher reaction temperature (350℃-400℃) and with longer reaction time (20-30 min) [38].
In 2005, I. Repins, C. Wolden and H. Ullal used plasma-assisted co-evaporation of S and Se for wide band gap chalcopyrite photovoltaic.
In this paper, the Se molecular species (Sen, 2<n<8) were converted to atomic species by low-pressure inductively-coupled plasma (ICP). The Gibbs energy of formation at 300K of Se atomic species was higher than Se molecular. So the atomic species required less energy to exceed the barrier of the formation of intermediate. In traditional fabrication of high-quality CIGS films like co-evaporation process required high
23
substrate temperatures (>500℃ ), limiting the selection of substrate material. Therefore, this process expected to reduce the substrate temperature in the deposition films process and substrate material had more choices such as foil or stainless steel. This paper demonstrated with plasma activation the film was converted to the CIS chalcopyrite and no secondary phases were observed with substrate temperature Ts=320℃ [39].
In 2007, Shogo Ishizuka, etc, report on the growth of CIGS films using a RF-cracked Se-radical beam source. The ICP plasma dissociated Se vapor to smaller Se-radical which was high reactivity. High reactivity of active Se-radical species created the modification of the kinetics of film growth, enhances migration during growth and improved the ability of In-Ga interdiffusion. The above factors leaded to CIGS film changed for higher dense, smoother surfaces and larger grain size than without RF-cracked Se-radical beam source. Though the film quality was significant improvement, but the photovoltaic performance was no significant influence. The energy conversion efficiency of solar cell was reached 17.5%. The radical source growth CIGS film has been to reduce the substrate temperature and increase the utilization of selenium material [40].
In 2011, Shogo Ishizuka, etc, then they announced another focus of using a RF-plasma cracked Se-radical beam source. Radical-Se grown CIGS solar cells showed some different properties of Evaporation-Se grown CIGS solar cells. The conversion efficiencies of solar cells be enhance by the RF-plasma cracked Se-radical beam source due to
24
increase open circuit voltage and fill factor, though simultaneously decreased the short circuit current density. The sodium content of the CIGS film by radical selenium treatment was higher and the sodium expected to inhibit element interdiffusion. Therefore, film growth slows down created beneficial selenium diffusion in the CIGS film. After this process CIGS film has a better quality and caused to improve the open-circuit voltage and fill factor of the CIGS solar cell. The reason of decrease short-circuit current was the smooth surface morphology which created reduction of light trapping [41].
2.3 Prepared to Cu(In,Ga)Se
2Absorber Layer
A wide variety of thin-film deposition methods has been used to deposit Cu(InGa)Se2 thin films. In this paper, we categorize them in co-evaporation and sequential processes (selenization). And then the selenization processes include elemental Se atmosphere and stacked elemental layer. Following the exposition will discuss them in detail.
2.3.1 Co-evaporation Process
The co-evaporation process is the vacuum technique. The setup is show in Figure 2-1. The system has four Knudsen type MBE cells that allow an optimal control of the deposition speed (1-5Å /s) for every material and on the thickness for the deposited film (0.5-4μm). The base pressure in the vacuum chamber is carried out reaches a value of 10-6 Torr.
This technique should have high purity material Cu (Kurt Lesker, 99.999%), Ga (Alfa Aesar, 99.9999%). Se (Kurt Lesker, 99.999%) and In (Kurt Lesker, 99.999%). Source temperatures of every material are:
25
Tso(Cu)=1300-1400ºC, Tso(In) = 1000-1100ºC, Tso(Ga) = 1150-1250ºC and Tso(Se) = 300-350ºC. The process requires a substrate temerpature between 300ºC and 550ºC for a certain time during film growth.
The fabrication procedure of three step co-evaporation process has several species. The popular species is NREL three-step co-evaporation and conversion efficiency as high as 19.5% [42]. At first In, Ga and Se are evaporated with different rates and deposited as (In,Ga)2Se3 at 300°C on the substrate. Afterwards Cu and Se are evaporated and deposited on the substrate at elevated temperatures. At last In, Ga, and Se are evaporated again. The inverted three-stage process leads to smoother film morphology and high efficiency solar cells.
The generally properties of co-evaporation process are described following:
(1) Highest conversion efficiency of device.
(2) In expensive raw materials.
(3) Demonstrated in production is easier than other technique.
(4) The process control is difficult due to the substrate temperature is difference at different procedure.
(5) Material utilization is disappointing, because the most material is coating on the chamber.
(6) The cost per unit cell is high due to the material utilization is bad and maintain the vacuum system.
(7) The area of solar cell device is limited by evaporation length.
Because the chamber become large in order to make large area of solar cell device.
26
2.3.2 Prepare Precursor Layer by Sputtering
I order to grow precursor layer the schematic diagram of RF sputter system is show in Figure 2-2. After evacuation of chamberto 1*106 Torr, pure Ar is introduced into chamber to sputter target and maintain pressure of 1*102 Torr. The purity Ar with flow rate of 6-15 sccm is applied. Prior to deposition of thin films, pre-sputtering is done for 10 min to clean contamination under working pressure of 3*102 Pa.
The sputtering technique is much the same and the only thing that difference of the target. In recent years the research studies in this technique that change many target or stack element for the purpose of get higher conversion efficiency.
The Cu-In-Ga ternary sputtering target was manufactured by turning a ternary-alloyed ingot cast from a molten alloy melted by a vacuum arc-refining furnace. The source materials of the ingot had a pre-determined composition of high-purity ( 99.999 at %) constituents, Cu, In, and Ga, at 50, 35, and 15 at%, respectively. Use the RF-sputtering to deposit Cu-In-Ga layer on the Mo/glass. Then, a Se film was deposited by vacuum evaporation onto the Cu–In–Ga precursor layer (without
In the first step, two types of precursor structures were employed: (1) a CuInGa/Mo single layer and (2) a CuGa/CuInGa/Mo layer. In the first
27
type of structure (1), a 650 nm thick film of CuInGa/Mo precursors was deposited by DC magnetron sputtering of a Cu0.9In0.75Ga0.25 ternary target onto Mo-coated soda-lime glass substrates. In the second type of structure (2), an additional 85-nm thick CuGa film was sputtered on top of the CuInGa/Mo structure using Cu0.7Ga0.3 target. In the second step, the above mentioned metallic precursors were selenized using a Se vapor in a quartz tube furnace. The temperature of selenization process was 500℃
for 20 min in order to form the p-type CuInGaSe2 chalcopyrite structure [44].
The generally properties of sputtering process are described following:
(1) The process control is easy.
(2) Have high materials utilization due to the diffusion length of sputtering system is smaller than co-evaporation.
(3) This technique easily scalable to large-area manufacturing processes and more uniform than other technique for the large-area manufacturing process.
(4) High production rates.
(5) The conversion efficiency of CIGS solar cells by sputtering have remained low as compared to co-evaporation
2.3.3 Prepare Precursor Layer by Electro-Deposition
In recent years, there are many new technological developments to make the large area of CIGS solar cells. Electro-deposition has the potential to develop into the manufacturing technology of large area.
28
CIGS films were prepared adopting co-electro-deposition of the four elements of Cu, In, Ga and Se. The potentiostatic technique with a conventional three-electrode configuration was used, where the reference electrode was a saturated calomel electrode (SCE), the counter electrode was Pt mesh and the working electrode was a Mo-coated soda-lime glass substrate. The electro-deposition bath consisted of CuCl2, GaCl3, InCl3, H2SeO3 and 1.0 M Na-citrate. To obtain ideal CIGS thin films, InCl3 and GaCl3 in sufficient quantity were added to the chemical bath to adjust the final composition of CIGS near to ideal stoichiometry. The pH of the chemical bath was adjusted to be 1.5 by adding drops of concentrated HCl. The selenization of the films was carried out in a tubular furnace.
CIGS thin films were annealed at 550℃ in selenium atmosphere for 1 h.
To prevent Se atom of CIGS thin films escaping, excessive selenium powder as a selenium source instead of high toxic H2Se gas was added during the heat treatment [45].
The generally properties of electro-deposition process are described following:
(1) The process control is easy.
(2) Have high materials utilization due to electro-deposition use solution to deposition thin film.
(3) This technique easily scalable to large-area manufacturing processes.
(4) High production rate due to high deposition speed.
(5) The ability of uniform and adhesion need to be improved.
(6) The conversion efficiency of CIGS solar cells remained low.
29
2.4 Atmospheric pressure plasma system 2.4.1 Arc Plasma
The operation of arc plasma is similar to an arc-welding machine, where an electrical arc is struck between two electrodes. The high energy of arc creates high temperatures ranging from 3000℃ to 7000℃. The plasma is highly ionized gas which is enclosed in a chamber. The waste material is fed into the chamber and the intense heat of the plasma break down organic molecules into their elemental atoms. In the strict control of the process, these atoms recombine into harmless gases, such as carbon dioxide. Solids such as glass and metals are melted to form materials, similar to hardened lava, in which toxic metals are encapsulated. There is no burning or incineration and no formation of ash with plasma arc technology. There are two main types of plasma arc processes: plasma arc melter and plasma torch.
Plasma arc melters have very high destruction efficiency. They are very robust; they can treat any waste with minimal or no pretreatment;
and they produce a stable waste form. The arc melter uses carbon electrodes to strike an arc in a bath of molten slag. The consumable 28 carbon electrodes are continuously inserted into the chamber, eliminating the need to shut down for electrode replacement or maintenance. The high temperatures produced by the arc convert the organic waste into light organics and primary elements.
Combustible gas is cleaned in the off-gas system and oxidized to CO2
and H2O in ceramic bed oxidizers. Due to the use of electrical heating in the absence of free oxygen, the potential for air pollution is low. The
30
inorganic portion of the waste is retained in a stable, leach-resistant slag.
In plasma torch systems, an arc is struck between a copper electrode and either a bath of molten slag or another electrode of opposite polarity. Plasma torch systems have very high destruction efficiency with plasma arc systems; they are very robust; and they can treat any waste or medium with minimal or no pre-treatment. The inorganic portion of the waste is retained in a stable, leach-resistant slag. The air pollution control system is larger than for the plasma arc system, due to the need to stabilize torch gas.
2.4.2 Atmospheric Pressure Plasma Jet
Atmospheric pressure plasma jet is meaning that operating at atmospheric pressure and it is non-thermal glow discharge plasma system.
The non-thermal plasma generates highly reactive ions, electrons and free radicals. The reactive species are directed onto a surface where the desired chemistry occurs. However the overall gas temperature remains quite cold, but the electrons are quite hot, typically 50-300℃.
2.4.3 Corona Discharge
A corona is a process by which a current develops between two high-potential electrodes in air, by ionizing that fluid to create a plasma around one electrode, and by using the ions generated in plasma processes as the charge carriers to the other electrode.
Corona discharge usually involves two asymmetric electrodes, one highly curved such as the tip of a needle or a narrow wire, and another
31
one of low curvature such as a plate or the ground. The high curvature assures a high potential gradient around one electrode, for the generation of the plasma.
Coronas may be positive, or negative. This is calculated by the polarity of the voltage on the high curvature electrode. If the curved electrode is positive associated to the flat electrode, it will have a positive corona, and vice visa. The physics of positive and negative coronas are obviously different. This asymmetry structure is a result of the great difference in mass between electrons and positively charged ions, and so only the electron having the ability to undergo a significant degree of ionizing inelastic collision at common temperatures and pressures.
2.4.4 Dielectric Barrier Discharge (DBD)
Dielectric barrier discharges involve a specific class of high voltage, ac, gaseous discharges that typically operate in the near atmospheric pressure range. Their defining feature is the presence of dielectric layers that make it impossible for charges generated in the gas to reach the conducting electrode surfaces. With each half cycle of the driving oscillation, the voltage applied across the gas exceeds that required for breakdown, and the formation of narrow discharge filaments initiates the conduction of electrons toward the more positive electrode. As charge accumulates on the dielectric layer at the end of each filament, the
Dielectric barrier discharges involve a specific class of high voltage, ac, gaseous discharges that typically operate in the near atmospheric pressure range. Their defining feature is the presence of dielectric layers that make it impossible for charges generated in the gas to reach the conducting electrode surfaces. With each half cycle of the driving oscillation, the voltage applied across the gas exceeds that required for breakdown, and the formation of narrow discharge filaments initiates the conduction of electrons toward the more positive electrode. As charge accumulates on the dielectric layer at the end of each filament, the