**2. Band gap engineering in ZnO**

Band gap engineering is the process of controlling or altering the band gap of a material by controlling the composition of certain semiconductor alloys. It is well known that tailoring of the energy band gap in semiconductors by band-gap engineering is important to create barrier layers and quantum wells with matching material properties, such as lattice con‐ stants, electron affinity for heterostructure device fabrication [2, 3].

Band-gap engineering in ZnO can be achieved by alloying with MgO, CdO or BeO. The en‐ ergy band gap Eg(x) of ternary semiconductor AxZn1-xO (A = Mg, Cd or Be) can be calculat‐ ed by the following equation:

$$\text{Eg}(\mathbf{x}) = (1 - \mathbf{x}) \to\_{\text{ZnO}} + \mathbf{x}\_{\text{EAO}} - \text{bx} \ (1 - \mathbf{x}) \tag{1}$$

turned from 3.3 to 4.5eV with the increase of Mg contents from 0 to 0.5. Tampo et al investi‐ gated excitonic optical transition in a Zn1−*<sup>x</sup>*Mg*x*O alloy grown by radical source molecular beam epitaxy [12]. The strong reflectance peaks at room temperature were detected from 3.42eV (*x*=0.05) to 4.62eV (*x* = 0.61) from ZnMgO layers at room temperature. PL spectra at room temperature were also observed for energies up to 4.06eV (*x* = 0.44). Wassner et al studied the optical and structural properties of MgZnO films with Mg contents between *x* = 0 and *x* = 0.37 grown on sapphire by plasma assisted molecular beam epitaxy using a MgO/ ZnMgO buffer layer [13]. In their experiments, the *a*-lattice parameter was independent from the Mg concentration, whereas the *c*-lattice parameter decreases from 5.20Å for *x* = 0 to 5.17Å for *x* = 0.37, indicating pseudomorphic growth. The peak position of the band edge

Makino et al investigated the structure and optical properties of CdxZn1-xO films grown on sapphire (0001) and ScAlMgO4 substrates by PLD [14]. The band gap of CdxZn1-xO films was

by incorporating Cd2+ with Cd concentration of 7%. Both lattice parameters a and c increase with the increasing Cd content in ZnO, which was agreement with the larger atomic size of Cd compared with Zn. CdxZn1-xO films were also prepared on c-plane sapphires by metalorganic vapor-phase epitaxy. The fundamental band gap was narrowed up to 300meV for a maximum Cd concentration of ~5%, introducing a lattice mismatch of only 0.5% with re‐ spect to binary ZnO. Lai et al prepared the CdxZn1-xO alloy by conventional solid-state reac‐ tion over the composition range and found that CdO effectively decreased the electronic

**Figure 5.** Room temperature PL spectra of ZnO/Zn0.9Mg0.1O SQW with different well width. From Ref. [17].

Using MgZnO as barrier layers, Chauveau et al prepared the nonpolar a-plane (Zn,Mg)O/ZnO quantum wells (QWs) grown by molecular beam epitaxy on r plane sapphire and a plane ZnO substrates [16]. They observed the excitonic transitions were strongly blue-shifted due to the

. The band gap narrowing to 2.99eV was achieved

ZnO-Based Light-Emitting Diodes http://dx.doi.org/10.5772/51181 29

luminescence blue shifted up to 4.11eV for *x* = 0.37.

bandgap both in the bulk and near the surface ZnO [15].

estimated by Eg(y) = 3.29 − 4.40y + 5.93y2

where *b* is the bowing parameter and EAO and EZnO are the band-gap energies of compounds AO and ZnO, respectively. While adding Mg or Be to ZnO results in an increase in band gap, and adding Cd leads to a decrease in band gap [3, 8].

Both MgO and CdO have the rock-salt structure, which is not the same as the ZnO wurtzite structure. When Mg and Cd contents in ZnO are high, phase separation may be detected, while BeO and ZnO share the same wurtzite structure and phase separation is not observed in BeZnO [2, 8]. Ryu et al studied the band gap of BeZnO and did not observed any phase separation when Be content was varied over the range from 0 to 100mol%. Figure4 shows the *a* lattice parameter as a function of room-temperature *Eg* values in AxZn1-xO alloy. There‐ fore, theoretically, the energy band gap of AxZn1-xO can be continuously modulated from 0.9eV (CdO) to 10.6eV (BeO) by changing the A concentration [8]. Han et al reported the band gap energy of the BexZn1-xO can be tailored from 3.30eV (x = 0) to 4.13eV (x = 0.28) by alloying ZnO with BeO [9].

**Figure 4.** Energy band gaps, lattice constants and crystal structures of selected II-VI compounds. From Ref. [9].

Ohtomo et al investigated the band gap of MgxZn1-xO films grown on sapphire by PLD, where *x* is the atomic fraction [10]. The band gap of MgxZn1-xO could be increased to 3.99eV at room temperature as the content of Mg was increased upward to *x* = 0.33. Above 33%, the phase segregation of MgO impurity was observed from the wurtzite MgZnO lattice. Takagi et al reported the growth of wurtzite MgZnO film with Mg concentration of 51% on sap‐ phire by molecular-beam epitaxy [11]. The band gap energy of MgxZn1-xO was successfully turned from 3.3 to 4.5eV with the increase of Mg contents from 0 to 0.5. Tampo et al investi‐ gated excitonic optical transition in a Zn1−*<sup>x</sup>*Mg*x*O alloy grown by radical source molecular beam epitaxy [12]. The strong reflectance peaks at room temperature were detected from 3.42eV (*x*=0.05) to 4.62eV (*x* = 0.61) from ZnMgO layers at room temperature. PL spectra at room temperature were also observed for energies up to 4.06eV (*x* = 0.44). Wassner et al studied the optical and structural properties of MgZnO films with Mg contents between *x* = 0 and *x* = 0.37 grown on sapphire by plasma assisted molecular beam epitaxy using a MgO/ ZnMgO buffer layer [13]. In their experiments, the *a*-lattice parameter was independent from the Mg concentration, whereas the *c*-lattice parameter decreases from 5.20Å for *x* = 0 to 5.17Å for *x* = 0.37, indicating pseudomorphic growth. The peak position of the band edge luminescence blue shifted up to 4.11eV for *x* = 0.37.

Band-gap engineering in ZnO can be achieved by alloying with MgO, CdO or BeO. The en‐ ergy band gap Eg(x) of ternary semiconductor AxZn1-xO (A = Mg, Cd or Be) can be calculat‐

where *b* is the bowing parameter and EAO and EZnO are the band-gap energies of compounds AO and ZnO, respectively. While adding Mg or Be to ZnO results in an increase in band

Both MgO and CdO have the rock-salt structure, which is not the same as the ZnO wurtzite structure. When Mg and Cd contents in ZnO are high, phase separation may be detected, while BeO and ZnO share the same wurtzite structure and phase separation is not observed in BeZnO [2, 8]. Ryu et al studied the band gap of BeZnO and did not observed any phase separation when Be content was varied over the range from 0 to 100mol%. Figure4 shows the *a* lattice parameter as a function of room-temperature *Eg* values in AxZn1-xO alloy. There‐ fore, theoretically, the energy band gap of AxZn1-xO can be continuously modulated from 0.9eV (CdO) to 10.6eV (BeO) by changing the A concentration [8]. Han et al reported the band gap energy of the BexZn1-xO can be tailored from 3.30eV (x = 0) to 4.13eV (x = 0.28) by

**Figure 4.** Energy band gaps, lattice constants and crystal structures of selected II-VI compounds. From Ref. [9].

Ohtomo et al investigated the band gap of MgxZn1-xO films grown on sapphire by PLD, where *x* is the atomic fraction [10]. The band gap of MgxZn1-xO could be increased to 3.99eV at room temperature as the content of Mg was increased upward to *x* = 0.33. Above 33%, the phase segregation of MgO impurity was observed from the wurtzite MgZnO lattice. Takagi et al reported the growth of wurtzite MgZnO film with Mg concentration of 51% on sap‐ phire by molecular-beam epitaxy [11]. The band gap energy of MgxZn1-xO was successfully

Eg(x) = (1− x) EZnO + xEAO − bx (1 − x) (1)

ed by the following equation:

28 Optoelectronics - Advanced Materials and Devices

alloying ZnO with BeO [9].

gap, and adding Cd leads to a decrease in band gap [3, 8].

Makino et al investigated the structure and optical properties of CdxZn1-xO films grown on sapphire (0001) and ScAlMgO4 substrates by PLD [14]. The band gap of CdxZn1-xO films was estimated by Eg(y) = 3.29 − 4.40y + 5.93y2 . The band gap narrowing to 2.99eV was achieved by incorporating Cd2+ with Cd concentration of 7%. Both lattice parameters a and c increase with the increasing Cd content in ZnO, which was agreement with the larger atomic size of Cd compared with Zn. CdxZn1-xO films were also prepared on c-plane sapphires by metalorganic vapor-phase epitaxy. The fundamental band gap was narrowed up to 300meV for a maximum Cd concentration of ~5%, introducing a lattice mismatch of only 0.5% with re‐ spect to binary ZnO. Lai et al prepared the CdxZn1-xO alloy by conventional solid-state reac‐ tion over the composition range and found that CdO effectively decreased the electronic bandgap both in the bulk and near the surface ZnO [15].

**Figure 5.** Room temperature PL spectra of ZnO/Zn0.9Mg0.1O SQW with different well width. From Ref. [17].

Using MgZnO as barrier layers, Chauveau et al prepared the nonpolar a-plane (Zn,Mg)O/ZnO quantum wells (QWs) grown by molecular beam epitaxy on r plane sapphire and a plane ZnO substrates [16]. They observed the excitonic transitions were strongly blue-shifted due to the anisotropic strain state in heteroepitaxial QW and the reduction of structural defects and the improvement of surface morphology were correlated with a strong enhancement of the photo‐ luminescence properties. Su et al investigated the optical properties of ZnO/ZnMgO single quantum well (SQW) prepared by plasma-assisted molecular beam epitaxy [17]. The photolu‐ minescence peak of the SQW shifted from 3.31 to 3.37eV as the well layer thickness was de‐ creased from 6 to 2nm (Figure5). ZnO/MgZnO superlattices were also fabricated by laser molecular-beam epitaxy and the excitonic stimulated emission up to 373K was observed in the superlattices. The emission energy could be tuned between 3.2 and 3.4eV, depending on the well thickness and/or the Mg content in the barrier layers.

glass by magnetron sputtering and found that a carrier concentration exhibited only a

Wang et al studied the properties of In-doped ZnO crystal by the hydrothermal technique [23]. The indium-doped ZnO crystals have a resistivity lower than 0.015Ωcm with a free car‐ rier concentration (mostly due to indium donors) of 1.09×1019/cm3 at room temperature. Quang et al reported the In-doped ZnO films grown by hydrothermal [24]. The films had a

/Vs and a concentration of 6.7×1018 - 3.2×1019/cm3

/ Vs [26]. Chikoidze et al grew Cl-doped ZnO films by MOCVD with a

) and/or hydrogen incorporation. Considerable efforts have been

VII elements such as F and Cl are also used as n-type dopants in ZnO, which substi‐ tuted oxygen ions. Cao et al reported F-doped ZnO grown by PLD with a minimum resistivity of 4.83×10-4 Ωcm, with a carrier concentration of 5.43×1020cm-3 and a mobili‐

To realize ZnO-based LEDs, the most important issue is the fabrication of high quality ptype ZnO. However, undoped ZnO exhibits n-type conduction and the reliable p-type dop‐ ing of the materials remains a major challenge because of the self-compensation from native

made to obtain p-type ZnO by doping different elements (N, P, As, Sb, Li, Na and K) with various techniques [2, 3]. Here, we present the typical results of p-type ZnO materials.

Among all potential p-type dopants for ZnO, N is considered the most promising dopant due to similar ionic radius compared with oxygen. It substitutes O sites in ZnO structure, resulting in the shallow acceptors. N2, NO, N2O, NH3 and Zn3N2 are acted as N sources de‐ pended on growth techniques [2, 3]. Liu et al reported p-type ZnO:N films grown on c-sap‐ phire by plasma-assisted molecular beam epitaxy [28]. The anomalous Raman mode at 275cm-1 was confirmed to be related to substitution of N for O site (NO) in ZnO. The films

vestigated p-type ZnO films prepared on a-plane (11–20) sapphire by MOCVD [29]. The op‐ timized result was achieved at the temperature of 400°C with a resistivity of 1.72Ωcm, a Hall

type ZnO films by oxidation of Zn3N2 films grown by direct current magnetron sputtering

type N,Ga-codoped ZnO films prepared by sputtering ZnO:Ga2O3 target in N2O ambient

Beside N, other group V elements (P, As and Sb) are also used to be acceptor dopants to ob‐ tain p-type ZnO. However, first-principle calculations show that XO (PO, AsO and SbO) are deep acceptors and have high acceptor-ionization energies, owing to their large ionic radii as compared to O, which make it impossible for XO to dop ZnO efficiently p-type [32]. We

/Vs, and a hole concentration of 2.29×1018cm−3. Wang et al prepared p-

exhibited a hole concentration of 2.21×1016cm-3 and a mobility of 1.33cm2

[30]. For oxidation temperature between 350 and 5000

with a hole concentration of 5.78×1017cm-3 at 5000

[31]. The film deposited on sapphire at 5500

. The In-doped ZnO

ZnO-Based Light-Emitting Diodes http://dx.doi.org/10.5772/51181 31

/Vs. Zeng et al in‐

C, p-type ZnO:N films were achieved,

C. Kumar et al reported on the growth of p-

C exhibited p-type conduction with a hole con‐

were grown by sol–gel method [25].

slight change with the thickness variations [22].

films with a carrier concentration of 3.22×1020/cm3

mobility of 4.18-20.9cm2

resistivity of 3.6×10-3 Ω cm [27].

donor defects (Vo and Zni

mobility of 1.59cm2

centration of 3.9×1017 cm−3.

ty of 23.8cm2

**3.2. p-type ZnO**
