Chapter 1 Introduction
1.1 Basic Concept of TFT-LCD
1.1.1 History of TFT-LCD
Liquid crystals were discovered by the Austrian botanist Rheinitzer in 1888. Liquid crystal is a term that indicates the status of a substance that is neither solid nor liquid, e.g. soapy water. In 1963, Williams, while working for RCA discovered that the way light passes through liquid crystal changes when it is stimulated by an electrical charge. Five years later, another RCA researcher named Heilmeyer and his colleagues made a display prototype that applied this concept. This prototype's success marked the beginning of modern liquid crystal display (LCD) technology.
In the beginning, liquid crystals were initially too unstable to use as a material for manufacturing display units, resulting in several market problems for LCD technology until a professor of the University of Hull in the U.K. discovered a stable liquid crystal material (biphenyl).
In recent years, the growth of flat panel displays in many fields is apparent. The flat panel displays includes liquid crystal display (LCD), light emitting diode (LED), electroluminescent (EL) panel, vacuum fluorescent display (VFD), and plasma display (PDP). Liquid crystal displays (LCDs) are the most popular among the various flat panel displays. Of all these systems, the liquid crystal display is considered to be the most promising. In addition to being thin and lightweight, these displays run on voltages so low they can be driven directly by a Large Scale Integrated Circuit (LSI). Since LCDs consume low power, they can run for long periods on batteries. However, limited viewing angle characteristics, large response times and limitations in the electro-optic characteristics were some of the drawbacks of LCD for large information content displays like video displays and televisions.
Many researchers and engineers have been engaged in improving the characteristics of liquid crystal displays since that time. The first AMLCD was successfully demonstrated by Brody in 1973, where a CdSe TFT was used as a switching element for each pixel of a 120 x 120 matrix [1]. After LeComber reported the first amorphous hydrogenated silicon (a-Si:H) TFT [2], many groups started the development of AMLCD using a-Si:H TFTs formed on glass substrate. Color filters were used in the transmission mode to realize the color displays. AMLCDs currently are mainly available on the market which use a-Si:H TFTs and polycrystalline silicon (poly-Si) TFTs.
There are two trends in the development of TFT-LCD. One is the
with larger screen sizes and higher pixel-content are required. The other is toward compact LCD modules with high pixel density for use in cell phones, projection television sets and digital cameras as monitor display.
Amorphous hydrogenated silicon (a-Si:H) TFT-LCDs have been widely used for many applications in direct view displays such as notebooks, televisions, personal computers, and various monitors. LCD screen size has rapidly increased during the last 10 years. Small-size LCDs with high pixel densities are also used as light valves for LCD projectors and viewfinders for digital cameras. Poly-Si TFTs are suitable for such applications since the monolithic drivers around the screen preclude a limit on connector pitches between driver ICs and display signal lines.
Because LCD has unique and superior features such as compactness, thin, light and low-power consumption in principle, it is a promising technology for realizing high quality interface such as displays, projectors, and portable personal devices.
1.1.2 Principles of LCD Technology
Light passes through liquid crystals, following the direction in which the molecules are arranged. When the molecule arrangement is twisted 90 degrees as shown in Fig. 1-1, the light also twists 90 degrees as it passes through the liquid crystals. The molecules in liquid crystals are easily rearranged by applying voltage. When voltage is applied, molecules rearrange themselves along with the electric field vertically and light passes straight through along the arrangement of molecules. A combination of polarizing filters and twisted liquid crystal creates a liquid crystal display. When two polarizing filters (polarizers) are arranged along perpendicular polarizing axes, light entering from below is re-directed 90 degrees along the helix arrangement of the liquid crystal molecules so that it passes through the upper polarizer. The liquid crystal molecules straighten out and stop redirecting the angle of the light as voltage is applied. Therefore, no lights can pass through that area of the LCD and it makes that area darker than the surrounding areas.
The color filter of a TFT-LCD consists of three primary colors (Red, Green, and Blue). The elements of this color filter line up one-to-one within the unit pixels on the TFT-array substrate. Each pixel is divided into three subpixels, where one set of RGB subpixels is equal to one pixel. Because the subpixels are too small to distinguish separately, the RGB elements seem to the human eye as a mixture of the three colors.
It will produce lots of colors by mixing these three primary colors (RGB) (see Fig. 1-2).
Polarizer Color Filter
Color Filter Polarizer
Polarizer Backlight TFT Liquid Crystal Polarizer Backlight TFT Liquid Crystal AC Voltage
Fig. 1-1: Liquid crystal works by applying voltage.
Fig. 1-2: LCD panel produces colors by adjusting the proportion of the 3 color pixels.
Today's TFT-LCD has a sandwich-like structure (see Fig. 1-3). It consists of liquid crystal, two polarizers and glass substrates. The top substrate is color filter, the bottom substrate is TFT array. Two Polarizers control the light entering and leaving. Glass substrate stops the filtering of electricity from electrodes. Color is expressed through the use of R, G and B color filters. Transparent electrodes drive the TFT-LCD. A highly transparent material (ITO) is used that will not interfere with the quality of the image's integrity. Liquid crystal material is injected between two glass plates. TFT array control the liquid crystal direction by applying voltage. The display operates in the transmission mode with fluorescent lamp behind the panel.
Fig. 1-3: TFT-LCD has a sandwich-like structure.
Polarizer
Glass substrate
Backlight
The driving method of LCD is dividing roughly into passive matrix LCD and active matrix LCD.
Passive Matrix LCD (PMLCD)
In this method, the transparent electrodes are set on X and Y electrodes as shown in Fig. 1-4(a). There are no switching devices in passive matrix LCD (PMLCD) as shown in Fig. 1-4(b), and each pixel is addressed for more than one frame time. Because all electrodes are electrically connected together, the effective applied voltage must average the signal voltage pulses over several frame times. Therefore, this driving method results in a slow response time and a reduction of the maximum contrast ratio. The addressing of a PMLCD also produces a kind of crosstalk that produces fuzzy images because non-selected pixels are driven through a secondary signal-voltage path.
Passive matrix LCDs are used in watches, calculators, or cheap hand held video games. As the pixel density (resolution) increased, however, passive matrix technology was no longer suitable as crosstalk became an issue with displays larger than 100×100 pixels.
Active Matrix LCD (AMLCD)
The configuration of a typical AMLCD is shown in Fig. 1-5(a). The pixels are arranged in X-Y matrix formed on the bottom glass substrate as shown in Fig. 1-5(b).
(b) Circuitry (a) Structure
Fig. 1-4: (a) Structure of passive matrix drive system [3].
(b) Circuitry of passive matrix drive system [3].
(b) Circuitry (a) Structure
Fig. 1-5: (a) structure of active matrix drive systems [3].
(b) Circuitry of active matrix drive systems [3].
Each pixel has an a-Si:H TFT which operates as an analogue switch to control the stored charge in an LC capacitor. The capacitance is defined between the pixel electrode on the bottom substrate and a common electrode on the top substrate. Active-pixel technology was invented so that the voltage applied at each pixel was well isolated from the other pixels [4]. The TFT made a suitable low-conductivity switch as turned off, had sufficient conductivity in the on-state such that it could charge a liquid crystal pixel's capacitance, and polarize the liquid crystal.
Fig. 1-6: A schematic diagram of TFT-LCD array with controllers, power supply, and driver circuits shows the driving of an LCD panel.
VRAM board
Fig. 1-6 is a schematic diagram of the TFT-LCD array with controllers, a power supply, and other driver circuits shows the driving of an LCD panel. The TFT substrate consists of a TFT array and an array of external terminals on which large-scale integrations (LSIs) are bonded to drive the TFT panel. These LSIs are directly bonded to the glass with tape automated bonding (TAB) connectors, and they provide each pixel of the panel with video signals to drive the LCD panel.
1.2 Motivation and Objective
Hydrogenated amorphous silicon (a-Si:H) is used extensively in Thin Film Transistors (TFTs) for Flat Panel Displays (FPDs) and large area imagers, and it is also a promising photovoltaic material. a-Si:H TFTs have a low off-current and sufficient on-current for most applications. However, a-Si:H has poor carrier mobility. The low mobility of a-Si:H results in a limitation on pixel sizes for display and imaging applications. Poly-Si has been proposed as an alternative to a-Si:H, as it can have mobility of up to 300 cm2 V-1 s-1. High performance silicon devices on insulators have recently been incorporated in various applications, such as lightweight flat panel displays.
The key for realizing such systems (SOP) is low temperature polycrystalline silicon (LTPS) TFT. Low-thermal-budget techniques, such as plasma-assisted hydrogen-[5], metal-[6], and laser-induced crystallization [7-14] are commonly employed to LTPS as the channel regions. In particular, efficient absorption of ultraviolet (UV) laser irradiation by hydrogenated amorphous silicon (a-Si:H) results in high-quality polycrystalline silicon (poly-Si) by excimer laser annealing (ELA) [9-11].
Unlike annealing using continuous-wave [14] and long pulse lasers (tens ns range), nonlinear photon absorption and nonequilibrium thermodynamics are expected to dominate the interactions by the femtosecond laser pulses [15-20]. Such a nonlinear process provides precise and low fluence associated with femtosecond laser ablation [17-20]. Our group had presented a near-infrared femtosecond laser (λ =
800 nm) annealing with 50 fs pulse duration [21]. In order to investigate the femtosecond laser-induced crystallization of amorphous silicon further, we utilize blue femtosecond laser (λ = 400 nm) to do annealing experiment in this thesis. Then we study the structural characteristics of recrystallized amorphous silicon films. At last we make comparisons with near-infrared femtosecond laser annealing and excimer laser annealing.
We report blue femtosecond laser-induced crystallization of amorphous silicon with high crystallinity, using ultralow laser energy.
1.3 Organization of this thesis
In chapter 1 of this thesis, an overview of TFT-LCD technology is presented first. History of TFT-LCD and principles of LCD technology would be introduced in section §1.1. Then the motivation and objective of the thesis would be written in section §1.2.
In the beginning of chapter 2, low temperature poly-silicon (LTPS) is given in detail. In section §2.2, several recrystallization methods which include solid phase crystallization (SPC), excimer laser annealing (ELA), and femtosecond laser annealing (FLA) are introduced.
In chapter 3, our sample preparation is mentioned. After that, a brief introduction to our femtosecond laser pulse and the frequency doubling process separately are described. Then we could see the experimental components completely in the schematic diagram of experiment setup in section §3.3. In section §3.4, the characterizations of blue femtosecond laser annealing samples which include scanning electron microscope (SEM) images and Raman spectra are presented finally.
In chapter 4, our group’s near-infrared femtosecond laser annealing work is reviewed first. Comparisons between the differences of absorption coefficient and penetration depth of near-infrared femtosecond laser annealing and blue femtosecond laser annealing are discussed in section §4.1. Then, Comparisons between difference of mechanism of excimer laser annealing (ELA) and blue femtosecond laser annealing are reported in section §4.2.
Finally in chapter 5, summary of the blue femtosecond laser
annealing would be written in section §5.1 and suggestions to the future work are also presented in section §5.2.
Chapter 2
Properties and Preparation of Low Temperature Poly-Silicon
In the beginning of chapter 2, low temperature poly-silicon (LTPS) is given in detail. In section §2.2, several recrystallization methods which include solid phase crystallization (SPC), excimer laser annealing (ELA), and femtosecond laser annealing (FLA) are introduced
2.1 Introduction of Low Temperature Poly-Silicon
Low temperature poly-silicon (LTPS) technology has been studied for more than a decade, for the purpose of driver integration at the periphery of AMLCD. With increasing the display area and pixel density of TFT-LCD, high mobility TFTs are required for pixel driver of TFT-LCD in order to shorten the charging time of pixel electrodes.
However, it is very difficult because hydrogenated amorphous silicon (a-Si:H) has low mobility (0.5 - 1 cm2 V-1 s-1). The problem of the low carrier mobility for a-Si TFTs can be overcome easily by introducing poly-silicon film instead of a-Si as a channel layer of TFTs. The poly-Si is the most promising material for obtaining such high mobility TFTs for pixel drivers, moreover, the peripheral driver circuits can be integrated on the same substrate. Fig. 2-1 shows the crystal structure and electrical characteristics performance comparison between the silicon thin films.
From the diagrams (Crystal Structure) in Fig. 2-1, we could understand
the difference between amorphous silicon, poly-silicon, and single crystal silicon. The amorphous silicon is the non-crystalline form of silicon.
Silicon is normally bonded to four neighboring silicon atoms. This is also the case in amorphous silicon. However, it does not form a continuous crystalline lattice as in crystalline silicon. Some atoms may actually have dangling bonds, which occur when it does not bond to four neighboring atoms. Since not all the atoms are four-fold coordinated, amorphous silicon is said to be under-coordinated. These dangling bonds are defects in the continuous random network, which can be passivated by introducing hydrogen into the silicon. It then becomes hydrogenated amorphous silicon.
Because of the low mobility of the amorphous silicon, it’s not suitable for fast switch. Another series of diagrams also shows the mobility relationship between the three kinds of silicon (a-Si, poly-Si, and single crystal Si). If there were many barriers on the road, we could not drive fast. On the contrary, we could drive fast on the flat way. It is the same with the electrons move in the three kinds of silicon. If the electrons need to pass through the grain boundaries within the active-channel region of a TFT, the mobility is low. That’s why the poly-silicon’s mobility is higher than amorphous silicon, but still lower than single crystal silicon. Although the mobility of single crystal silicon is the highest, the conventional process temperature to manufacture the single crystal silicon film is over 1000 °C. But the melting point of the glass is about 550 °C. That’s why we cannot employ the single crystal silicon in TFT-LCD process.
0.002~0.004
Fig. 2-1: The crystal structure and electrical characteristics performance comparison between the silicon thin films.
From Fig. 2-1, we know the leakage current of poly-Si is larger than a-Si:H. Many groups recently have studied the fabrication of high performance poly-Si TFTs with low leakage current, high mobility, high on/off current ratio, and good uniformity. In order to create the high quality TFT-LCD without flicker and cross-talk, we need to utilize poly-Si TFT with low leakage current (below 1 pA / µm) [24].
Fig. 2-2 shows the concept of circuit integration in TFT-LCD. As carrier mobility is improved, higher extent of integration will be possible.
Low temperature poly-silicon allows integration of driver ICs onto the substrate. Both the number of external connections and the external size
of the substrate could be minimized. System on panel (SOP) technology, implementing memories, sensors, and controllers with driver circuits of displays on glass, will be realized in the future with poly-Si thin-film transistors (TFTs) [22].
Fig. 2-2: Concepts of circuit integration. The comparisons of a-Si TFT-LCD, poly-Si TFT-LCD, and poly-Si TFT-LCD system on panel (SOP).
We could see another advantage of poly-silicon in Fig. 2-3. The transistor of LTPS LCD is smaller than that of amorphous-silicon. Thus,
the aperture ratio of poly-silicon TFT-LCD and its screen brightness could be increased significantly [23].
Fig. 2-3: The difference of aperture ratio between a-Si TFT and poly-Si TFT [23].
The most conventional method to fabricate poly-Si is low pressure chemical vapor deposition (LPCVD) [24]. However, this method have some disadvantages such as high deposition temperature over 600°C, small grain size < 50 nm, poor crystallinity and high grain boundary states. Another disadvantage is the cheap glass substrate used in low temperature and large area processes could not be used in high temperature process. To enhance crystal properties, SPC is more useful method to increase the grain size than the as-deposited poly-Si film by LPCVD [25]. But it needs a long time annealing at high temperature over 600 °C. Now, there is a liquid phase recrystallization method such as ELA, which enable us to have good quality poly-Si film with large grain size and good electrical properties. ELA technology can be fabricate the
low temperature poly-Si TFT-LCDs without mechanical stress of glass substrate. Laser-induced crystallization is a promising process for obtaining a high quality poly-Si films [9-14]. But the excimer laser consists of toxic gas. People need to be careful as using excimer laser especially.
Several recrystallization methods to produce poly-silicon would be explained in detail in the next paragraph which include solid phase crystallization (SPC), excimer laser annealing (ELA), and femtosecond laser annealing (FLA)
2.2 Preparation of Poly-Silicon
2.2.1 Solid Phase Crystallization （SPC）
Solid phase crystallization (SPC) of amorphous silicon attracted many interests for the application to poly-Si TFT-LCDs [29]. It is the most direct method of obtaining poly-Si films, from initially amorphous silicon films, is via SPC in furnace environment. Amorphous silicon is a thermodynamically metastable phase, processing a driving force for transformation to polycrystalline phase given sufficient energy to overcome the initial energy barrier.
Unfortunately, SPC using low-temperature furnace annealing requires very long anneal times and hence suffers from a substantial tradeoff between performance and throughput [27]. The crystallization from an amorphous phase to a polycrystalline phase occurs through two processes—nucleation and grain growth [27]. Both of them have characteristic activation energies. The nucleation activation energy is extracted from the time to onset of crystallization, while the grain growth rate is extracted from grain progression data. The nucleation activation energy is larger than the grain growth activation energy for the SPC system [28]. To achieve the largest possible grains, it is desirable to suppress nucleation relative to grain growth. Therefore, SPC is typically done at a low temperature. This results in a reduction in throughput through an increase in the incubation time and a decrease in the grain growth rate. Higher temperatures increase throughput oppositely.
However, excessive nucleation results in smaller grains and hence poorer performance.
Thus SPC can be accomplished within a wide annealing temperature range that requires a similar wide range of annealing time (i.e.
time required for complete transformation of the precursor-Si film to poly-Si). The relationship between annealing temperature and annealing time is not unique. Based on the above transformation method, large grain size relates to longer crystallization time. For practical applications, the crystallization time corresponding to average grain size exceeding 0.5-1 µm may be prohibitingly long. The typical SPC poly-Si microstructure is characterized by a large density of structural defects.
The result of this high grain-defect density is a saturation in the electrical performance of poly-Si TFTs, fabricated with such poly-Si films with grain size larger than approximately 0.3-0.5 µm. Therefore, standard SPC technology can only produce poly-Si TFTs of mediocre performance.
This translated to a mobility range 20 - 40 cm2 V-1 s-1 and the threshold voltage ranges from 3 to 6 V [11].
Despite of the mediocre performance, solid phase crystallization (SPC) is a promising technique due to its simplicity, low cost, and
Despite of the mediocre performance, solid phase crystallization (SPC) is a promising technique due to its simplicity, low cost, and