**2. Graphene and graphene field-effect transistors**

polarizer, and so on. Among all these desired THz components and devices, modulators stand at on the focus of current interest [5]. Modulators can be used to control the amplitude, phase, polarization state, spatial propagation direction, pulse shape, and many more characteristic properties of electromagnetic waves and thus act as essential parts of sophisticated

However, unlike the optical or microwave regime where active modulators are well established, the THz frequency regime is still in great demand for efficient, fast, and versatile active light modulators [6]. This is mainly due to the lack of natural materials that have tunable electromagnetic response to THz wave [5]. Nevertheless, materials including semiconductors [7, 8], metamaterials [9–12], superconductors [13, 14], and phase-transition materials [15, 16] have been intensively explored to control and manipulate THz wave with great progress being made in this direction. Modulators can be categorized by the technique or material system, which is employed to modulate the wave, for example, optical, electronic, thermal, and magnetic modulators.

The all-optical approach is an effective and attractive method to fabricate THz modulators, especially those with broadband or spatial operating features. Semiconductors, especially high-resistive (HR) silicon, have proven to be suitable for all-optical modulation of THz wave by converting photons into electrons upon optical illumination [17]. Generally, the pump laser produces a temporary region of high absorption or reflectance on semiconductor. THz wave, co-projected on this area, is thus modulated [18]. As a result, all optical spatial THz modulators, based on a bare silicon wafer, have previously been proposed to realize photodesigned THz devices [19] or reconfigurable quasi-optical THz components [20, 21]. Si is an attractive candidate for optical opponents because it is earth abundant, chemically stable, and has a suitable band gap. Silicon-based THz devices are particularly desirable as they would enable interfacing with existing and emerging Si-based optoelectronics, thereby providing potential low-cost route toward applications. However, a silicon wafer exhibits strong reflection to both optical light (~40%) and THz radiation (~30%), which greatly limits the achievable tunability and versatility. Xie et al. verified that the modulation depth (MD) of silicon wafer to THz wave is only 19.9% under 800 mW femtosecond laser, although this value could be 98.6% if the pump laser intensity is high enough [21]. Very recently, plasmonic layers [22], graphene [23, 24], and even thin organic layers [25] have been fabricated on the surface of Si to enhance the modulation properties. Si modulators with these additional layers can work under lower pumping power, while having the same or even two to four times larger modulation depth. Very recently, Qi-Ye Wen and his co-workers demonstrated an interesting Si nanostructure for optically driven THz modulators [26]. They showed that nanotip (SiNT) arrays made from silicon wafer can be utilized as antireflection layers for both THz wave and visible light to achieve a low-loss and spectrally broadband THz modulator with a remarkably enhanced MD. Instead of fabricating heterogeneous materials on silicon, the nanotips are directly etched from the Si substrate and thus are structurally stable. Compared with the modulators fabricated on bare silicon, a nearly three times larger MD is achieved with the SiNT modulator. Crucially, the intrinsic THz transmission of the SiNT modulator is as high as 90% due to a strong antireflection effect arising from the nanotip layer as a result of the

One disadvantage of the semiconductor-based all-optical THz modulators is its relatively low modulation speed. The ultimate modulation speed is decided by the carrier recombination

formation of graded refractive index on the Si surface.

THz application systems such as wireless telecommunication or security imaging [6].

118 Design, Simulation and Construction of Field Effect Transistors

Graphene is a quasi-two-dimensional isolated monolayer of carbon atoms that are arranged in a hexagonal lattice. It is well known for its remarkable electron mobility at room temperature, with reported values in excess of 15,000 cm2 ·V−1·s−1 [34]. Hole and electron mobilities were expected to be nearly identical. Graphene holds great promise for various material/ device applications, including solar cells [35], light-emitting diodes (LED), touch panels, and smart windows or phonesal [36].

Graphene is a zero-band-gap semiconductor where conduction and valence bands meet at the Dirac points. The band gap is an extremely important characteristic of the semiconductor for transistor application, which enables the transistor device to turn off and minimizes leakage current at off state. In this circumstance, the gapless band structure of single-layer graphene makes it unsuitable for the direct use of graphene- based field-effect transistors (FET), though it is one of the most widely discussed applications in electronics. In order to overcome the challenges faced in incorporating graphene into microelectronic applications, great efforts have gone into developing three main aspects including developing a synthesis technique to manufacture graphene over wafer-scale areas, forming a high-quality gate dielectric on the surface of graphene, and, most importantly, opening an energy band gap in graphene. This results in the observation that narrow ribbon widths down to less than 10 nm are required to open a band gap in graphene and achieve an acceptable level of low off currents. However, fabricating of narrow ribbons of that dimension is a big challenge even with current advanced lithographic techniques [37, 38].

The zero-band-gap characteristic means that the density of states (DOS) in graphene is linear with respect to the energy level. Therefore, gate voltage can modulate the DOS linearly to enable modulation of carrier (current) in the channel. In other words, graphene-based fieldeffect transistor (GFET) can be used to tune the carrier concentration in graphene by applying a voltage at the gate, making it possible to modulate the absorption/transmission of THz wave through the devices. Recently, such a graphene-based THz modulator was demonstrated by Sensale-Rodriguez et al. and Maeng et al. [39, 40]. It is reported that the transmission of THz wave through graphene can be controlled by electrically tuning the density of states available for intra-band transitions. Though the modulation depth and speed are limited to 15% and 20 kHz by this electrical device prototype, it opens a new direction for graphene application in a transistor structure [39, 40].

Large-area monolayers of graphene films were synthesized by chemical vapor deposition (CVD)

characterized by a Raman spectroscopy with a 514 nm laser in **Figure 2(a)**. The two obvious peaks

half-maximum (FWHM) of the two-dimensional peak is about 38 cm−1. In addition, the intensity of

To acquire further understanding of the THz modulation characteristics, we now discuss the carrier properties of the graphene layer under electrical biasing. As graphene can be treated as a thin film, the frequency-dependent amplitude transmission |T(ω)| is given by ∣ *T*(ω) ∣ =

where Γ is the carrier-scattering rate and *D* is the Drude weight. *D* can be further expressed as

constant, and *n* is the carrier concentration of graphene. The corresponding variation of the *EF*

rier concentration *n.* **Figure 2(b)** shows the conical band structures of graphene, whose Fermi level and carrier concentration can be changed by applying different gate biases in a GFET. As a result, the transmittance of the THz wave could be modulated by varying the applied bias voltages at the gate. Ambient atmosphere and processing residual on the surface of graphene often deviates the Dirac point of graphene from the zero voltage point [47]. To tune the transmittance of

 is the effective refractive index of the substrate. σ (ω) is the complex sheet conductivity of graphene, which can be described by the simple Drude model, namely σ(ω) <sup>=</sup> \_\_\_\_\_\_\_ *iD*

O3

confirming that the device surface is smooth, uniform, and free of pinholes over large areas.

O3

*D/IG* ratio of ~0.18, indicating that the transferred graphene

O3

/p-Si and SiO2

O3


film on the Si substrate. The inset shows the surface of graphene/Al2

is the Fermi velocity, *e* is the electron charge, ℏ is the reduced Planck

clearly shows the relation between *EF*

/p-Si substrate was used to fabricate the refer-

Graphene Field-Effect Transistor for Terahertz Modulation

http://dx.doi.org/10.5772/intechopen.76744

/p-Si substrate are the G peak at ~1591 cm−1 and

/p-Si substrate and 300-nm

*G/I2D* is ~0.43, and the full-width at

/p-Si substrate were

O3

121

/p-Si substrate [45,

π(ω + iΓ) ,

and the car-

in the inset of **Figure 1(b)**,

is the vacuum impedance,

on a copper foil, which were then transferred onto the ~60-nm Al2

O3

/p-Si substrate, respectively [43, 44]. The SiO<sup>2</sup>

obtained by optical microscope. The dash lines are a guide to the eye.

O3

in the Raman spectra of graphene on the Al2

, where *VF*

in graphene extracted from |*EF*<sup>|</sup> <sup>=</sup> <sup>ℏ</sup> *VF*

the D peak at 1341 cm−1 is low with an *I*

ence sample. The transferred graphene monolayers on the Al2

two-dimensional peak at ~2687 cm−1. The peak intensity ratio *I*

46]. The optical microscope shows the top view of the graphene/Al2

is a high-quality monolayer which is similar to the sample grown on the SiO2

) <sup>∣</sup>, where *<sup>N</sup>* <sup>=</sup> 1 is the number of graphene layer, Z0

(*π*|*n*|)1/2

SiO<sup>2</sup>

**Figure 1.** (a) θ–2θ scan of Al2

section of a single crystalline ~60 nm Al2

<sup>∣</sup> \_\_\_\_\_\_\_\_\_\_\_\_\_\_ <sup>1</sup> <sup>1</sup> <sup>+</sup> *<sup>N</sup>* <sup>Z</sup><sup>0</sup> σ(ω)/(1 <sup>+</sup> *ns*

/ℏ) (*π*|*n*|)1/2

and ns

*<sup>D</sup>* <sup>=</sup> (*VF <sup>e</sup>*<sup>2</sup>
