Effects of microwave annealing on electrical enhancement of amorphous oxide
semiconductor thin film transistor
Li-Feng Teng, Po-Tsun Liu, Yuan-Jou Lo, and Yao-Jen Lee
Citation: Applied Physics Letters 101, 132901 (2012); doi: 10.1063/1.4754627 View online: http://dx.doi.org/10.1063/1.4754627
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/101/13?ver=pdfcov Published by the AIP Publishing
Articles you may be interested in
Coplanar amorphous-indium-gallium-zinc-oxide thin film transistor with He plasma treated heavily doped layer Appl. Phys. Lett. 104, 022115 (2014); 10.1063/1.4862320
Effect of annealing time on bias stress and light-induced instabilities in amorphous indium–gallium–zinc-oxide thin-film transistors
J. Appl. Phys. 110, 114503 (2011); 10.1063/1.3662869
Trap-limited and percolation conduction mechanisms in amorphous oxide semiconductor thin film transistors Appl. Phys. Lett. 98, 203508 (2011); 10.1063/1.3589371
Correlation of photoconductivity response of amorphous In–Ga–Zn–O films with transistor performance using microwave photoconductivity decay method
Appl. Phys. Lett. 98, 102107 (2011); 10.1063/1.3561755
Role of order and disorder on the electronic performances of oxide semiconductor thin film transistors J. Appl. Phys. 101, 044505 (2007); 10.1063/1.2495754
Effects of microwave annealing on electrical enhancement of amorphous
oxide semiconductor thin film transistor
Li-Feng Teng,1Po-Tsun Liu,2,a)Yuan-Jou Lo,3and Yao-Jen Lee4
1
Department of Photonics and Institute of Electro-Optical Engineering, National Chiao Tung University, Hsinchu 30010, Taiwan
2
Department of Photonics and Display Institute, National Chiao Tung University, HsinChu 30010, Taiwan
3
College of Photonics and Institute of Photonic System, National Chiao Tung University, Tainan City 71150, Taiwan
4
National Nano Device Laboratories, Hsinchu 30078, Taiwan
(Received 13 August 2012; accepted 10 September 2012; published online 24 September 2012) By using microwave annealing technology instead of thermal furnace annealing, this work elucidates the electrical characteristics of amorphous InGaZnO thin film transistor (a-IGZO TFT) with a carrier mobility of 13.5 cm2/Vs, threshold voltage of 3.28 V, and subthreshold swing of 0.43 V/decade. This TFT performance with microwave annealing of 100 s is well competitive with its counterpart with furnace annealing at 450C for 1 h. A physical mechanism for the electrical improvement is also deduced. Owing to its low thermal budget and selective heating to materials of interest, microwave annealing is highly promising for amorphous oxide in semiconductor TFT manufacturing.VC 2012 American Institute of Physics. [http://dx.doi.org/10.1063/1.4754627]
Owing to their high carrier mobility, transparency for visi-ble light, high process compatibility with present solid-state semiconductor technologies, and available room-temperature deposition, transparent amorphous InGaZnO thin films have received considerable attention for their use in next-generation active-matrix liquid-crystal display (AMLCD) and organic light-emitting diode display (AMOLED) technologies.1–5 Sputter deposition technology is widely adopted, especially for the large-size amorphous InGaZnO thin film transistor (a-IGZO TFT) display backplane manufacturing. The sputter-deposited a-IGZO active layer typically requires thermal annealing at around 400C for 30 min or longer to achieve a satisfactory device performance and stability.6–8For conven-tional furnace annealing, thermal energy is transferred to the material of interest by creating a temperature gradient from outward to inward, resulting in additional thermal energy con-sumption. However, furnace annealing is limited by low effi-ciency of heating and high thermal budget process. This work presents a novel microwave annealing process for a-IGZO TFT fabrication with low thermal budget process. Microwave heating process can transfer the energy directly to the target materials by absorption of microwave energy throughout the volume of the material. Among its advantages include thermal uniformity, rapid heating process, shortened manufacturing period, low thermal budget, and suppression of unexpected species diffusion.9Microwave heating is also characterized by selective heating of materials, which is impossible with the typical furnace annealing process.10,11This work also investi-gates how microwave annealing affects the electrical perform-ance and reliability of a-IGZO TFTs. A physical mechanism responsible for the electrical enhancement of a-IGZO TFT is also proposed by comparing with the counterpart of a-IGZO TFT thermally annealed in a typical furnace at 450C for 1 h.
TFT devices with an inverted staggered bottom gate structure without backchannel passivation were fabricated on
a glass substrate. A 100-nm-thick Mo layer was formed as a bottom gate electrode in a DC sputtering system, followed by the subsequent deposition of a 150-nm-thick silicon nitride (SiNx) on the patterned gate electrode by
plasma-enhanced chemical vapor deposition (PECVD). Next, the active channel layer of a 50-nm-thick a-IGZO layer was formed by DC sputtering at a power of 100 W with an argon flow rate of 10 sccm at room temperature. The target for the a-IGZO film deposition was an IGZO pellet with the compo-nent ratio of 1:1:1:4 at. % (In:Ga:Zn:O). A 100-nm-thick in-dium tin oxide (ITO) was then formed to act as source/drain electrodes by RF sputtering; all of the layers were defined by the shadow masks. The channel width and length of a-IGZO TFTs were varied from 200 to 1000 lm. All samples were annealed sequentially in a microwave annealing system with a microwave frequency of 5.8 GHz,12as shown in Fig.1(a). The microwave power used in this work is 600 W and 1200 W, as denoted by 1P and 2P, respectively. Microwave annealing lasted from 100 s to 600 s. The absorption of elec-tromagnetic wave energy is generally in the frequency range of 2 to 18 GHz for ZnO-based materials.13,14Therefore, as is expected, the a-IGZO active layer effectively absorbs the microwave energy with the frequency we used in this work. Additionally, a control sample of a-IGZO TFT with a typical thermal annealing process was fabricated for comparison with microwave annealed a-IGZO TFT. The thermal anneal-ing was conducted in a furnace with N2 gas flow rate of
10 l/h at 450C for 1 h. Electrical and reliability measure-ments were carried out using the semiconductor parameter analyzer, Keithley 4200. For the x-ray photoelectron spec-troscopy (XPS) measurement, 50-nm a-IGZO thin films with various microwave annealing conditions were deposited separately on n-type Si wafer.
Figure1(b)shows the transfer characteristics of a-IGZO TFTs with different microwave annealing conditions at a drain-to-source voltage (VDS) of 11 V. The extracted threshold
voltage (Vth) and mobility of each a-IGZO TFT were also
a)
E-mail: [email protected].
compared in the inset. According to this figure, the threshold voltage of a-IGZO TFT decreased from 11.4 to 1.62 V, as the microwave annealing duration increased from 100 s to 600 s, respectively, with an identical microwave power. Moreover, the mobility increased when increasing the microwave anneal-ing time, indicatanneal-ing that increasanneal-ing the microwave annealanneal-ing time improved the electrical characteristics of a-IGZO TFTs. As microwave power increases from 1P to 2P with the same process time, the mobility of a-IGZO TFTs is enhanced signif-icantly from 4.86 to 13.9 cm2/Vs; in addition, the threshold voltage is lowered from 11.4 to 3.13 V.
This work also studied how gate bias stress (GBS) affects a-IGZO TFTs with different microwave annealing conditions, as shown in Fig. 1(c). In the gate bias stress measurements, an electric field of 2.5 MV/cm was applied to gate electrode and source/drain electrodes were grounded. This finding suggests that increasing both microwave power and treatment duration can improve the electrical reliability of a-IGZO TFT. The Vth shift of microwave annealed
a-IGZO TFT after GBS testing was decreased from the value of 16.2 V with the condition of 1P power for 100 s to a mini-mum of 1.6 V with 2P power for 100 s. Also, increasing the microwave power and annealing duration can enhance microwave energy absorption capacity for the a-IGZO active layer. The material analysis method was performed further to observe the evolution of a-IGZO film structure, allowing us to identify the physical mechanism responsible for improving the electrical properties of a-IGZO TFT devices.
Figure2shows the XPS analysis results of O1s spectrum for the a-IGZO thin film with microwave annealing treat-ments. Two components of O1s peaks could be fitted by Gaussian Lorentzian deconvolution, which centered at 530.6 and 531.4 eV, respectively. The lower binding energy cen-tered at 530.6 eV, denoted as peak A, originated from the lat-tice oxygen ions with neighboring metal atoms.15,16 The higher binding energy peak at 531.4 eV, denoted as peak B, corresponds to O2ion at an oxygen-deficient region in the
matrix of the a-IGZO film.8,17 According to the results of XPS, peak A increased and peak B decreased while the microwave annealing duration increased, as shown in Figs. 2(a)–2(c), respectively. Also, the XPS spectrum of a-IGZO film with high-power microwave annealing was com-posed of a high intensity of peak A and low intensity of peak B, as shown in Fig.2(d). This figure revealed a high content of lattice oxygen ions in the microwave-annealed a-IGZO films with few oxygen-deficient regions as the microwave power was increased. XPS analysis results confirmed that the microwave annealing process assisted the oxygen ions bind-ing with metal atoms and suppressed the formation of oxygen-deficient region in the a-IGZO films when increasing both the microwave power and annealing duration.
Figure 3 compares the effects of microwave and typi-cally used furnace annealing on a-IGZO TFT. The a-IGZO TFT with 2P microwave annealing for 100 s exhibited a superior performance with a higher mobility and lower sub-threshold swing than that of the 450C furnace-annealed a-IGZO TFT as shown in the inset of Fig.3(a). After GBS, the microwave annealed a-IGZO TFT also revealed a compara-ble reliability with the furnace-annealed one, as shown in Fig. 3. Trap densities (Nt) can be estimated from
subthres-hold swing value with the following formula:
S:S:¼ loge10 kBT=e½1 þ eðtNtþ DitÞ=Ci; (1)
where kBdenotes the Boltzmann constant; T represents the
temperature; e is the elementary electric charge; t is the thickness of the a-IGZO active layer; andDitis the interface
trap density, assuming the traptNtdominated andDitis
neg-ligible.6The Nt value of a-IGZO TFT with microwave and
furnace annealing was 2.49 1017 and 3.51 1017cm3,
respectively. The microwave-annealed a-IGZO exhibited a lower trap density than that of the furnace-annealed one. Figure 3(b) shows the XPS O1s spectrum of a-IGZO thin film with 2P microwave annealing for 100 s and furnace FIG. 1. (a) The schematic of microwave annealing system in this work. (b) The drain current (ID) versus gate voltage
(VG) curves of a-IGZO TFTs with
dif-ferent microwave annealing processes, at a drain-to-source voltage (VDS) of
11 V. The symbols of 1P_100s, 1P_300s, 1P_600s, and 2P_100s stand for the microwave annealed a-IGZO TFTs with 1P (600 W) for 100, 300, and 600 s, and 2P (1200 W) for 100 s, respec-tively. The corresponding device param-eters, such as threshold voltage (Vth) and
mobility of a-IGZO TFTs are shown in the inset. (c) Threshold voltage shift of microwave-annealed a-IGZO TFTs as a function of gate bias stress durations. The electric field of gate bias stress was fixed at 2.5 MV/cm. The Vthshift was
defined as the difference of the threshold voltages before and after the gate bias stress application to the a-IGZO TFT device.
annealing at 450C for 1 h for comparison. The O1s peak of a-IGZO film with microwave annealing composed of a higher intensity of peak A and a lower intensity of peak B than those of the 450C furnace-annealed one. These mate-rial analysis results were consistent with the electrical improvement in device performance and reliability of a-IGZO TFT, since microwave annealing facilitated the for-mation of lattice oxygen and eliminated the defects originat-ing from oxygen deficiency. Energy transfer to the a-IGZO TFT was even more effective by the microwave annealing process than that of the conventional furnace annealing.
In summary, this work has demonstrated the feasibility of high performance and reliable a-IGZO TFTs with micro-wave annealing process. Micromicro-wave annealing with low thermal budget can reduce the manufacturing process period and improve electrical characteristics of a-IGZO TFTs, due to the effective absorption of microwave energy by the a-IGZO active layer. This selective heating also potentially avoided the damage to materials neighboring the a-IGZO channel layer in the TFT device structure during thermal
processes. With optimum microwave annealing around 1200 W for 100 s in this work, electrical performance and reliability of a-IGZO TFT are more significantly promoted than with furnace annealing at 450C for 1 h. Results of this study significantly contribute to microwave annealing appli-cations for us in emerging flat-panel displays of transparent oxide TFTs technology.
The authors would like to thank the National Science Council of the Republic of China, Taiwan for financially supporting this research under Contract No. NSC 100-2628-E-009-016-MY3. National Nano Device Laboratories (Hsin-chu, Taiwan) was commended for the use of its facilities.
1
K. Nomura, H. Ohta, A. Takagi, T. Kamiya, M. Hirano, and H. Hosono,
Nature432, 488 (2004).
2J. S. Park, K. S. Kim, Y. G. Park, and Y. G. Mo,Adv. Mater.
21, 329 (2009).
3
C. J. Chiu, S. P. Chang, and S. J. Chang,IEEE Electron Device Lett.31, 1245 (2010).
4P. T. Liu, Y. T. Chou, L. F. Teng, and C. S. Fuh,Appl. Phys. Lett.
97, 083505 (2010).
FIG. 3. (a) The comparison of threshold voltage shift between the a-IGZO TFT with microwave annealing at 2P for 100 s and furnace annealing at 450C for
1 h, as a function of gate bias stress durations. The electric field of gate bias stress was fixed at 2.5 MV/cm. The inset shows threshold voltage (Vth),
subthres-hold swing (S.S), and mobility of the fresh a-IGZO TFT devices. (b) The comparison of XPS O1s spectrum for the a-IGZO films with the furnace annealing at 450C for 1 h (upper side) and the microwave annealing at 2P for 100 s (lower side).
FIG. 2. XPS analyses of O1s spectrum for the a-IGZO films with microwave annealing at (a) 1P for 100 s, (b) 1P for 300 s, (c) 1P for 600 s, and (d) 2P for 100 s, respectively. Peak A denotes the lower binding energy peak centered at 530.6 eV, generated from the lattice oxy-gen ions with neighboring metal atoms. Peak B is the higher binding energy peak at 531.4 eV corresponding to O2 ion at an oxygen-deficient region with the matrix of the a-IGZO film.
5P. T. Liu, Y. T. Chou, L. F. Teng, F. H. Li, and H. P. Shieh,Appl. Phys. Lett.98, 052102 (2011).
6
K. Nomura, T. Kamiya, H. Ohta, M. Hirano, and H. Hosono,Appl. Phys. Lett.93, 192107 (2008).
7K. Nomura, T. Kamiya, Y. Kikuchi, M. Hirano, and H. Hosono,Thin Solid Films518, 3012 (2010).
8
C. S. Fuh, S. M. Sze, P. T. Liu, L. F. Teng, and Y. T. Chou,Thin Solid Films520, 1489 (2011).
9
J. R. Groza, S. H. Risbud, and K. Yamazaki, J. Mater. Res.7, 2643 (1992).
10
T. F. Schulze, H. N. Beushausen, T. Hansmann, L. Korte, and B. Rech,
Appl. Phys. Lett.95, 182108 (2009).
11H. L. Hortensius, A. €Ozt€urk, P. Zeng, E. F. C. Driessen, and T. M.
Klapwijk,Appl. Phys. Lett.100, 223112 (2012).
12Y. J. Lee, F. K. Hsueh, S. C. Huang, J. M. Kowalski, J. E. Kowalski, A. T.
Y. Cheng, A. Koo, G. L. Luo, and C. Y. Wu,IEEE Electron Device Lett.
30, 123 (2009).
13Y. J. Chen, M. S. Cao, T. H. Wang, and Q. Wan,Appl. Phys. Lett.84,
3367 (2004).
14
R. F. Zhuo, L. Qiao, H. T. Feng, J. T. Chen, D. Yan, Z. G. Wu, and P. X. Yan,J. Appl. Phys.104, 094101 (2008).
15S. Y. Han, G. S. Herman, and C. H. Chang,J. Am. Chem. Soc.133, 5166
(2011).
16
H. Cho, E. A. Douglas, A. Scheurmann, B. P. Gila, V. Craciun, E. S. Lamber, E. S. Pearton, and F. Ren, Electrochem. Solid-State Lett.14, H431 (2011).
17T. Jun, K. Song, Y. Jeong, K. Woo, D. Kim, C. Bae, and J. Moon,J. Mater. Chem.21, 1102 (2011).
