Improved Performance of NILC Poly-Si Nanowire TFTs by Using Ni-Gettering
Bau-Ming Wanga, Tzu-Ming Yanga, YewChung Sermon Wua, Chun-Jung Sub, and Horng-Chih Linb
a
Department of Materials Science and Engineering, National Chiao Tung University, Hsinchu, Taiwan
b
Department of Electronics Engineering, National Chiao Tung University, Hsinchu, Taiwan
Nickel contamination inside nickel-metal induced lateral crystallization polycrystalline silicon is an issue to fabricate high performance nanowire thin film transistors. The phosphorous-doped -Si/chem-SiO2 films were employed as Ni-gettering layers
to investigate effect of Ni residues on the performance of NILC poly-Si NW TFTs. It was found that the performance of NW TFTs was greatly improved after Ni-gettering process.
Introduction
Low-temperature polycrystalline silicon (poly-Si) thin-film transistors (TFTs) have attracted considerable interest for their application in active-matrix liquid crystal displays (AMLCDs) on cheap glass substrate (1). Recently, high performance poly-Si nanowire (NW) TFTs have been fabricated by nickel-metal induced lateral crystallization (NILC) (2-3). Since NILC grain could be formed parallel to the channel direction, it becomes feasible to form Si NWs with nearly monocrystalline structures (4). Unfortunately, poly-Si/oxide interfaces and poly-Si grain boundaries trap Ni and NiSi2 precipitates, thus
increasing leakage current (5-8) and shifting the threshold voltage (9). Since the poly-Si/oxide interfaces/volume ratio of NW TFTs was much higher than that of tradition NILC TFTs, effect of Ni residues on the performance of NILC NW TFTs should be investigated. In this study, phosphorous-doped -Si/chem-SiO2 films were used to
investigate effect of Ni residues on the performance of NILC NW TFTs.
Experimental
Two kinds of NWs were investigated in this study. One was designated as “NILC NW” which was a poly-Si NW fabricated with traditional NILC methods, and another was “GETR NW” which utilized the same traditional NILC method with an additional Ni-gettering process. An approach for making NILC NW channels similar to Ref. (2) and (10) was followed. The detailed procedures were basically identical to those described in the Ref. (10), like gate dielectric deposition and source/drain formation. In this study, the major difference was NILC process performed before NW channel was defined. NILC length was about 17 Pm after lateral crystallization at 540o
C for 24 h.
As for GETR NW, an additional Ni-gettering process (11) was carried out to reduce the Ni concentration in NILC film. The gettering structure (phosphorous-doped -Si/chem-SiO2) is shown in Fig. 1(a). After gettering at 550°C for 12 h,
phosphorous-ECS Transactions, 33 (5) 169-172 (2010) 10.1149/1.3481233 © The Electrochemical Society
169 ecsdl.org/site/terms_use) unless CC License in place (see abstract). address. Redistribution subject to ECS terms of use (see
140.113.38.11
doped -Si and chem-SiO2 layers were removed using 5% tetra-methyl ammonium
hydroxide (TMAH) and 1% DHF solution, respectively. For comparison, NILC films were also subjected to an extended heat treatment as Ni-gettering condition. These two poly-Si films were then subjected to an anisotropic etching to form poly-Si spacers (NWs) in a self-aligned manner. The device structure is illustrated in Fig. 1(b).
Fig. 1 (a) Ni-gettering structure fabricated with phosphorous-doped -Si capped on the chem-SiO2. (b) The schematic of the proposed poly-Si NW TFT structure.
Results and Discussion
The TFT devices with a couple of NW channels have a nominal channel (L) of 0.8 Pm and an effective channel width (W) of 140 nm (2 × WNW). Typical ID-VG transfer
characteristics of NW TFTs at VD = 0.5 and 3 V are compared in Fig. 2. The measured
and extracted key device parameters are summarized in Table I. The threshold voltage (VTH) is defined at a normalized drain current of ID = (W/L) × 100 nA at VD = 0.5 V. The
subthreshold swing (S.S.) is extracted at VD = 0.5 V. The field-effect mobility (PFE) is
extracted from the maximum value of transconductance at VD = 0.5 V. The leakage
current (IOFF) is defined as the minimum drain current along the gate voltage at VD = 3 V.
Fig.2 ID-VG transfer characteristics of NILC poly-Si NW TFTs with and without
Ni-gettering.
ECS Transactions, 33 (5) 169-172 (2010)
170 ecsdl.org/site/terms_use) unless CC License in place (see abstract). address. Redistribution subject to ECS terms of use (see
140.113.38.11
As shown in TABLE I, ten NILC and GETR NW TFTs were measured. The performance of NILC NW TFTs was improved after Ni-gettering process. GETR NW TFTs had lower IOFF, higher ION/IOFF ratio, and higher PFE compared with NILC NW
TFTs. This improvement indicates the trap state density (Ntrap) was effectively reduced
using phosphorous-doped -Si gettering processes. The trap state density of the TFTs was extracted using Levinson and Proano’s method (12). The trap density of GETR NW TFTs is 2.52 × 1012 cm-2, which is less than that of NILC NW TFTs (3.95 × 1012 cm-2). Since GETR grains (boundaries) were similar to NILC grains, the reduction in Ntrap values
implies that those Ni-related defects have been reduced using phosphorous-doped -Si/chem-SiO2 gettering structure (11).
TABLE I. Device characteristics for ten NILC NW TFTs with and without Ni-gettering. Parameters (W/L=2x70nm/0.8m) GETR NW TFTs NILC NW TFTs IOFF (pA) 13.0 ± 5.4 81.9 ± 95.7 ION/IOFF ratio (10 6 ) 3.2 ± 1.0 1.4 ± 1.2 FE (cm2/V-s) 140.7 ± 38.2 117.3 ± 17.8 S.S. (mV/dec) 418 ± 63 395 ± 66 VTH (V) 0.35 ± 0.22 0.19 ± 0.40
Besides, as shown in TABLE I, the VTH of the NILC NW TFT is 0.19 V, which is
less than that of GETR NW TFT (0.35 V). This is because Ni residues could cause a high density of positive charge at the oxide/NILC poly-Si interface (13). The negative shift of VTH of NILC NW TFT was due to the presence of these positive charges and
nickel-related donor-like defects.
Summary
In this study, high performance NILC NW poly-Si TFTs with a couple of 70-nm NW channels were fabricated, and then improved by Ni-gettering process. The phosphorous-doped -Si/chem-SiO2 films were employed as Ni-gettering layers to reduce Ni residues
within NILC poly-Si film. After Ni-gettering process, the performance of NILC NW TFTs was improved. GETR NW TFTs had lower IOFF, higher ION/IOFF ratio, and higher
PFE compared with NILC NW TFTs.
Acknowledgments
This work was funded by Sino American Silicon Products Incorporation and the National Science Council of the Republic of China under Grant No. 98-2221-E-009 -041-MY3. Technical supports from the National Nano Device Laboratory, Center for Nano Science and Technology and the Nano Facility Center of the National Chiao Tung University are also acknowledged.
References
1. M. Stewart, R. S. Howell, L. Pires, and M. K. Hatalis, IEEE Trans. Electron Devices, 48, 845 (2001).
2. C. J. Su, H. C. Lin, and T. Y. Huang, IEEE Electron Device Lett., 27, 582 (2006).
ECS Transactions, 33 (5) 169-172 (2010)
171 ecsdl.org/site/terms_use) unless CC License in place (see abstract). address. Redistribution subject to ECS terms of use (see
140.113.38.11
3. C. W. Chang, S. F. Chen, C. L. Chang, C. K. Deng, J. J. Huang, and T. F. Lei, IEEE Electron Device Lett., 29, 474 (2008).
4. J. Gu, S. Y. Chou, N. Yao, H. Zandbergen, and J. K. Farrer, Appl. Phys. Lett., 81, 1104 (2002).
5. P. J. van der Zaag, M. A. Verhejen, S. Y. Yoon, and N. D. Young, Appl. Phys. Lett., 81, 3404 (2002).
6. G. A. Bhat, Z. Jin, H. S. Kwok, and M. Wong, IEEE Electron Device Lett., 20, 97 (1999).
7. Z. Jin, K. Moulding, H. S. Kowk, and M. Wong, IEEE Electron Device Lett., 20, 167 (1999).
8. D. Murley, N. Young, M. Trainor, and D. McCulloch, IEEE Trans. Electron
Devices, 48, 1145 (2001).
9. G. A. Bhat, H. S. Kwok, and M. Wong, Solid State Electron., 44, 1321 (2000). 10. H. C. Lin, M. H. Lee, C. J. Su, T. Y. Huang, C. C. Lee, and Y. S. Yang, IEEE
Electron Device Lett., 26, 643 (2005).
11. B. M. Wang, and Y. S. Wu, J. Electro. Mater., 38, 767 (2009).
12. R. E. Proano, R. S. Misage, and D. G. Ast, IEEE Trans. Electron Devices, 36, 1915 (1989).
13. Y. Lee, S. Bae, and S. J. Fonash, IEEE Electron Device Lett., 26, 900 (2005).
ECS Transactions, 33 (5) 169-172 (2010)
172 ecsdl.org/site/terms_use) unless CC License in place (see abstract). address. Redistribution subject to ECS terms of use (see
140.113.38.11