**3. Moore's law**

In the article published in April 1965, one of the cofounders of Intel Corporation, Dr. Gordon Earle Moore, predicted that the number of electronic components (which include not just transistors but capacitors, resistors, inductors, diodes, etc. as well) in an IC chip would double every year [9]. Ten years later, Moore revised his prediction to a doubling of every 2 years. Moore's prediction, which is more commonly known as Moore's law nowadays, has been widely used by the IC manufacturers as a tool to predict the increase of components in a chip for the coming generations [10]. To date, Moore's law has been proven to have held valid for close to half a century. **Table 1** tabulates the progressive trend of the integration level for the semiconductor industry. It can be observed from the table that the number of transistors that can be fabricated in a chip has been growing continuously over the years. In fact, this growth has been in close agreement with Moore's law. In order to highlight the technological advancement in the IC industries, each decade since the inception of the semiconductor transistor has been earmarked as a different era. Eight eras have existed hitherto—they are the small-scale integration (SSI), medium-scale integration (MSI), large-scale integration (LSI), very large-scale integration (VLSI), ultra-large-scale integration (ULSI), super largescale integration (SLSI), extra-large-scale integration (ELSI) and giant large-scale integration (GLSI) eras. During the VLSI era, a microprocessor was fabricated for the first time into a single IC chip. Although this era has now long passed, the VLSI term is still being commonly coined today. This is partly due to the absence of a significant qualitative leap between VLSI and its subsequent eras, and partly, it is also because IC engineers have been so used to this term; they decided to continue adopting it.


**7**

form the inversion layer.

*Introductory Chapter: Integrated Circuit Chip DOI: http://dx.doi.org/10.5772/intechopen.92818*

devices (such as microprocessors) in 2011.

**4.1 The MOSFET**

Today, the transistors fabricated in an IC chip are mostly MOSFETs. The earliest paper describing the operation principle of a MOSFET can be traced back to that reported in Julius Edgar Lilienfeld's patent in 1933 [11]. Unfortunately, the technology at that time was inadequate to allow Lilienfeld's idea to be physically materialized. In 1959, Dr. Dawon Kahng and Dr. Martin M. (John) Atalla at the BTL successfully constructed the MOSFET [12]. In 1963, two engineers from the Radio Corporation of America (RCA) Princeton laboratory, Dr. Steven R. Hofstein and Dr. Frederic P. Heiman, presented the theoretical description on the fundamental nature of the silicon planar MOSFET [13]. In the same year, Dr. Tom Chih-Tang Sah and Dr. Frank Marion Wanlass of Fairchild Semiconductor invented the first complementary metal oxide semiconductor (CMOS) logic circuit [14]. In 1989, Dr. Digh Hisamoto and his team member at Hitachi Central Research Laboratory introduced the fin field-effect transistor or better known as the FinFET—a nonplanar MOSFET modified from its planar counterpart. Although the FinFET was found to possess various advantages over the planar MOSFET, it was not adopted by the industries then. This was partly due to the difficulty in fabricating its threedimensional structure and, partly, also because the planar MOSFETs still had plenty of rooms to be improved further. Having realized that the planar MOSFET was gradually approaching its bottleneck in its technological advancement, chipmakers started to resort to FinFETs in the fabrication of high-end electronic

The MOSFET is nothing more than a device which operates as an electronic switch. **Figure 2** shows the basic structure of the MOSFET. The transistor comprises four terminals, namely, the drain (*D*), source (*S*), gate (*G*) and substrate or body (*B*) terminals. As can be clearly seen from the figure, the device constitutes three layers—a polysilicon layer (which forms the gate terminal), an oxide layer (known the gate oxide) and a single-crystal semiconductor layer (known as the substrate). In the early days, the gate terminal was made of aluminum. It is from these three layers of materials that the FET device acquired its name. In the mid-1970s, however, the gate material was replaced with polysilicon. When ion implantation was introduced to form the self-aligned source and drain terminals in the 1970s, a hightemperature (higher than 1000°C) annealing process was required to repair the damaged crystal structure at the surface of the substrate, as a result of the energetic dopant ion bombardment and to activate the dopant [15]. IC engineers observed that the aluminum gate melted during the annealing process. This is because aluminum has a melting point of about 660.3°C. In order to overcome this problem, polysilicon which has a melting point of about 1414°C was employed as the replacement for gate material. Although the gate today is no longer made of aluminum, the

term MOSFET has been so widely accepted that it stays until today.

The basic operation principle of a MOSFET is actually quite straightforward. When a voltage source is connected in between the drain and source terminals, a conducting channel is to be formed between the two terminals to allow the current to flow. The channel is commonly referred to as the inversion layer since the charges accumulated at the channel oppose those of the substrate. In this case, the gate terminal acts like a switch which controls the formation of the inversion layer. When sufficient voltage drop (and, of course, with the appropriate polarity) is applied to the gate terminal, carriers would be attracted to the gate oxide-substrate interface to

**4. The field-effect transistors**

#### **Table 1.**

*Integration level of an integrated circuit chip.*
