Effects of Ga₂O₃ deposition power on electrical properties of cosputtered In–Ga–Zn–O semiconductor films and thin-film transistors

IGZO films were deposited on SCHOTT B270 glass substrates at room temperature using In₂O₃ (purity, 99.99%), and Ga₂O₃ (purity, 99.99%) ceramic targets and a Zn (purity, 99.99%) metallic target, all 3 in. in diameter, as shown in Fig. 1(a). Details of the preparation of targets are found elsewhere.²⁵⁾ IGZO films were deposited using a two radio-frequency (RF) (ceramics targets: In₂O₃ and Ga₂O₃) and one direct-current (DC) (metal target: Zn) magnetron cosputtering system (LJ-UHV LJ-303CL) at room temperature, in which the Ga₂O₃ deposition power was varied, in order to verify the effects of Ga₂O₃ deposition power on the film properties of IGZO channels. The deposition powers of the In₂O₃ and Zn targets were fixed at 100 and 75 W, respectively. The thicknesses of the IGZO films under different deposition powers were controlled within 180–200 nm. The deposition rates under different deposition powers and the corresponding film thicknesses were measured using an n&k analyzer 1200. The deposition chamber was initially evacuated to 5.3 × 10⁻⁴ Pa, and a fixed Ar gas flow (50 sccm) and O₂ gas flow (10 sccm) were introduced into the chamber to maintain the working pressure at 0.67 Pa. After deposition, the samples were annealed at 400 °C at a working pressure of 0.67 Pa for 1 h in N₂ ambient of 40 sccm. The resistivity of these films was measured using a four-point probe (Napson RT-80). Hall measurements were performed by the van der Pauw method using a HALL8800 system (Swin). The magnetic field applied during the measurement was 0.68 T. The test samples were cut into 1 × 1 cm² squares. The surface morphology and chemical composition of the cosputtered IGZO films were respectively investigated using a scanning electron microscope (SEM; JOEL JSM 6500-F) at an operating voltage of 15 kV and an energy dispersive spectrometer (EDS; SEM-S4700). The crystallinity of the IGZO films was also analyzed using glancing angle X-ray diffraction (PANalytical X'Pert Pro) analysis with a Ni-filtered Cu Kα (λ = 1.5418 Å) source at a glancing incident angle of 1°. The scanning range was between 2θ = 20 and 80°.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The device structure is of the inverted-staggered type, which is the most commonly used structure for active matrix liquid crystal displays (AMLCDs), as shown in Fig. 1(b). To fabricate this structure, a 200 nm Al–Si–Cu film was first deposited by physical vapor deposition (PVD) on a 4 in. silicon substrate capped with a 500-nm-thick thermally grown silicon dioxide (SiO₂) film. The metal layer was patterned by photolithography and subsequent wet etching steps to form the gate electrode. Then, a 100 nm tetraethylorthosilicate (TEOS) oxide was deposited by plasma-enhanced chemical vapor deposition (PECVD) as the gate dielectric. Before depositing the IGZO active layer, we presputtered the target with argon flow for 15 min to clean the surface of each target. Subsequently, a 50 nm IGZO film was deposited as the channel layer using three-target cosputtering at various deposition powers of the Ga₂O₃ target. The deposition powers of the In₂O₃ and Zn targets were fixed at 100 and 75 W, respectively. The sputtering conditions were the same as the cosputtering condition for the IGZO films. After the deposition, the IGZO channel was postannealed using a backend vacuum annealing furnace at 300 °C at a working pressure of 6.7 Pa for 1 h in N₂ ambient of 40 sccm. A 300 nm Al–1.5 wt % Ti S/D metal was then formed by a lift-off process. Afterwards, a lithographic step for defining the active device region was performed. A diluted HCl solution $\text{(HCl}:\text{H}_{2}\text{O} = 1:200)$ was used instead to avoid damage and severe lateral etching of the IGZO channel film. In order to achieve contact with the gate electrode, contact etching was performed by wet etching using a buffer oxide etcher (BOE). The channel width (W) was fixed at 400 µm and the designed channel length (L), which is defined as the distance between the source and drain metal pads, was varied from 10 to 100 µm. The electrical measurement of all devices was executed using an Agilent 4156A precision semiconductor parameter analyzer, and the measurement temperature was maintained at 25 °C. Prior to the measurement, all the IGZO TFTs samples used in this study were annealed at 200 °C in air for 40 min on a hot plate to remove excess moisture on TFTs.

The deposition rate of co-sputtered IGZO films increase from 3.7 to 5.4 nm/min with the Ga₂O₃ deposition power. Figure 2 shows the Hall measurement plot of the cosputtered IGZO films as a function of Ga₂O₃ deposition power. The carrier concentration and Hall mobility clearly decrease as the Ga₂O₃ deposition power increases. Specifically, the carrier concentration decreases to less than 3 × 10¹⁶ cm⁻³ when the Ga₂O₃ power is 175 W and the Hall mobility decreases from 12.8 cm² V⁻¹ s⁻¹ and saturates at 4.6 cm² V⁻¹ s⁻¹ with increasing Ga₂O₃ deposition power. The results indicate that the film resistivity increases considerably with the Ga₂O₃ deposition power, owing to the lower carrier concentration and lower Hall mobility. However, the carrier concentration and Hall mobility show abnormal values when the Ga₂O₃ deposition power is 200 W because the film resistance is outside of the range of the Hall measurements. Hu and Gordon²⁶⁾ reported that when the solubility limit reachs 1.0 at. % Ga, the sheet resistance of Ga-doped ZnO films increased gradually. This is due to the reduction in the density of free charge carriers, and the interstitial occupation by Ga atoms, which leads to neutral defects in the structure without contributing a free electron.²⁷⁾

Zoom In Zoom Out Reset image size

Figure 3 shows SEM graphs of the cosputtered IGZO films with increasing Ga₂O₃ deposition power. It can be observed that the surface morphology shows smaller grains with increasing RF power of Ga₂O₃. From the SEM studies, a thin film deposited at a higher deposition power yielded smaller grains with an average size of about 10–20 nm as shown in Fig. 3(d) than those of its counterparts, as shown in Figs. 3(a)–3(c). This result is attributed to the increasing number of nucleation centers during the incorporation of the dopant into the host material.²⁶⁾ The reason for the Hall mobility decrease with increasing Ga₂O₃ deposition power correlates with the decrease in the grain size of IGZO films. Figure 4 shows the In, Ga, and Zn/(In + Ga + Zn) atomic ratios of cosputtered IGZO films as a function of Ga₂O₃ deposition power. The results were measured by the EDS technique. Although this was only a relative comparison between the conditions, it was clear that the at. % Ga of the deposited films could be controlled by adjusting the sputtering power for the Ga₂O₃ target. Figure 4 shows that the Zn/(In + Ga + Zn) ratio decreases clearly from 77.1 to 55.1%, whereas the Ga/(In + Ga + Zn) ratio of the deposited films increases noticeably from 8.4 to 28.9% and the In/(In + Ga + Zn) ratio increases slightly from 14.4 to 16% inside the IGZO films as a function of Ga₂O₃ deposition power. The result in Fig. 2 indicates film resistivity increases considerably with increasing Ga₂O₃ deposition power. Therefore, the zinc atoms increase the conductivity and gallium atoms enhance the resistivity of the cosputtered IGZO films. Thus, an appropriate addition of Ga is effective in suppressing oxygen vacancy formation and represents an effective way to attain lower carrier concentrations; similar findings have been reported by others researchers.^15,16) The results of the chemical composition analyses showed that Ga dopants inside the IGZO films inhibited the grain growth of the prepared films with increasing Ga₂O₃ power, as shown in Fig. 3.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

In order to verify the relationships between the crystalline phases and electrical properties of the cosputtered IGZO films, the crystallinities of the IGZO films with various Ga₂O₃ deposition powers were also analyzed using glancing angle X-ray diffraction; the result are shown in Fig. 5. The IGZO film cosputtered using three targets of In₂O₃, Ga₂O₃, and Zn reveals a polycrystalline oxide film. Two crystalline phases of InGaZn₇O₁₀ and InGaZn₂O₅ are shown in the studied films. With increasing Ga₂O₃ deposition power, the crystallinity of the InGaZn₇O₁₀ phase decreases, and the InGaZn₇O₁₀ phase is transformed to the InGaZn₂O₅ phase, which is ascribed to the fact that the $\text{In}:\text{Ga}:\text{Zn}$ ratio is varied from $14.5:8.5:77$ to $\text{16}:\text{29}:\text{55}$ as the Ga₂O₃ deposition power increases from 100 to 200 W, as shown in Fig. 4. The increase in the resistivity for the cosputtered films correlates with the decreasing crystallinity of the InGaZn₇O₁₀ phase and the phase transformation from InGaZn₇O₁₀ to InGaZn₂O₅ with increasing Ga₂O₃ deposition power.

Zoom In Zoom Out Reset image size

Figure 6 shows the transfer characteristics of the cosputtered IGZO TFTs at V_D = 10 V and W/L = 400 µm/10 µm with the deposition power of Ga₂O₃ varying from 100 to 200 W. The results indicate that the device performance shows an obviously low on-current (I_on)/off-current (I_off) ratio when the Ga₂O₃ power is 100 W, which is ascribed to a significantly high I_off (>10⁻⁴ A). However, the I_on/I_off ratio clearly increases to 1 × 10⁹ when the Ga₂O₃ power is increased to 150 W, owing to the fact that all the devices exhibit a sufficiently low I_off of less than 1 × 10⁻¹³ A as the Ga₂O₃ power is more than 150 W. In terms of the device off-current, compared with that of the IGZO channel prepared at 100 W deposition power, a significant improvement in the off-current (<10⁻¹³ A) has been achieved in IGZO TFTs with a channel prepared at a power of 150 W or higher, which is ascribed to the fact that the resistivity in the IGZO channels increases considerably from 0.035 to 0.653 Ω cm with the deposition power of Ga₂O₃ varying from 100 to 150 W, as shown in Fig. 2. In addition, oxygen vacancies are the main source of free electrons released for transport in the metal–oxide–semiconductors. This is clearly confirmed in the figure as the transfer curves shift positively with increasing Ga₂O₃ deposition power, since the Ga-rich film tends to suppress carrier generation and oxygen vacancy formation.^15,16) Therefore, a higher gate voltage is necessary to accumulate free electrons to form a conductive layer between the source and the drain. Table I shows a summary of the electrical characteristics of the cosputtered IGZO TFTs with various deposition powers of Ga₂O₃ and W/L = 400 µm/10 µm. The saturation mobility (μ_sat) was calculated using

$$\begin{equation} I_{\text{D}} = \frac{W}{2L}\mu_{\text{sat}}C_{\text{OX}}(V_{\text{G}} - V_{\text{th}})^{2}, \end{equation} \tag{ 1 }$$

where C_OX and V_th are the capacitance of the TEOS gate insulator and the threshold voltage, respectively. V_th is defined as the intercept voltage with V_G from the maximum slope of the respective square-root (sqrt) I_D vs V_G plot, as shown in Fig. 6. Figures 7(a) and 7(b) show the I_D–V_D output characteristics of the cosputtered IGZO TFTs (W/L = 400 µm/10 µm) with deposition powers of Ga₂O₃ of 150 and 200 W, respectively. With an optimum deposition power of Ga₂O₃ at 150 W, a higher saturated drain current (4.5 µA) at a higher drain voltage is demonstrated at V_G − V_th = 8 V, which is ascribed to the existence of sufficient oxygen vacancies for releasing free electrons for transport in cosputtered IGZO semiconductors. By contrast, the output characteristics of IGZO TFTs at a higher Ga₂O₃ power of 200 W indicate a lower unsaturated drain current at a higher drain, suggesting the lack of strong inversion in a high-resistance IGZO channel. For a low V_D, the total resistance (R_total) as a function of the designed channel length can be evaluated by the total resistance method conducted in the linear region of the output characteristics of the devices using^28–32)

$$\begin{equation} R_{\text{total}} = \frac{V_{\text{D}}}{I_{\text{D}}} = R_{\text{CH}} + R_{\text{SD}}, \end{equation} \tag{ 2 }$$

where R_CH and R_SD are the channel resistance and source/drain (S/D) parasitic resistance, respectively. Figures 8(a) and 8(b) show the extracted R_total, R_CH, and R_SD as a function of gate overdrive with L = 400 and 10 µm for the IGZO TFT devices with deposition Ga₂O₃ powers of 150 and 200 W, respectively. As can be seen in Fig. 8(a), for the lower Ga₂O₃ deposition power of 150 W, the R_SD (30 kΩ) contribution to R_total is negligible regardless of the gate overdrive voltage. On the other hand, for the devices prepared with a higher Ga₂O₃ deposition power of 200 W, the R_SD contribution to device characteristics is not negligible, as shown in Fig. 8(b). Actually R_SD (250 kΩ) even approaches R_CH at a higher gate overdrive voltage of 8 V, as shown in the figure. The devices prepared with a higher Ga₂O₃ deposition power of 200 W have a lower drain current at the same gate overdrive voltage owing to the higher R_SD, as shown in Fig. 7(b). In addition, Fig. 7(b) indicates the I_D–V_D output characteristics of the cosputtered IGZO TFTs with a higher Ga₂O₃ power of 200 W showing the decrease in the unsaturated drain current with increasing drain voltage due to lack of a strong inversion at a high V_D in the high-resistance IGZO channel. Therefore, the cosputtered IGZO TFT prepared with a Ga₂O₃ power of 200 W shows the worst device characteristics, as shown in Table I. It is clear that a higher gate overdrive voltage induces more free electrons to be transported in the channel, which is the main reason for the decrease in channel resistance. Because the cosputtered IGZO films deposited with a higher Ga₂O₃ power tend to reduce their carrier concentration in the channel, a higher potential barrier height and longer transport paths are expected from the percolation model.²⁾ By controlling the sputtering power of Ga₂O₃, polycrystalline cosputtered IGZO thin films have been successfully deposited, and the fabricated TFTs revealed good device performance.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size