高介電常數閘極介電層材料製備與可靠性分析

行政院國家科學委員會專題研究計畫 成果報告

高介電常數閘極介電層材料製備與可靠性分析

中 華 民 國 94 年 4 月 6 日

高介電常數閘極介電層材料製備與可靠性分析

Preparation and characterization of high-K gate dielectric materials

Abstract

HFO2 is a promising alternative for replacing SiO2 in ULSI manufacturing. In this work,

we employ different gas plasma, i.g. N2O, NH3, and N2, to nitrify the HfO2 films in order to

obtain less bulk traps responsible to threshold voltage instability. In this work, we found that

HfO2 film with N2O plasma nitridation exhibits excellent properties, such as lower swing,

higher conductance and driving current. This is due to the N2O plasma nitridation can reduce

interface states and bulk traps in the HfO2 film, confirmed by using charge pumping

technique1) with different measurement method, i.e. fixed base, fixed peak and fixed

amplitude method.

Introduction

Charge trapping is an important concern in the CMOS devices with HfO2 gate dielectrics,

due to the large amount of bulk traps presented in the HfO2 films. Their presence can result in

the reliability degradation,2) mobility degradation3-7) and threshold voltage instability8-12). In

order to improve the film quality, a variety of nitridation techniques were used to incorporate

nitrogen into the high-k films.4,9-11) In this work, we utilized the N2O plasma treatment

following the HfO2 deposition, which possesses the advantage of low thermal budget for

preventing the HfO2 films from crystallization during processing. It was found that the

post-deposition N2O plasma treatment can not only effectively improve the electrical

characteristics of the pMOSFETs with the HfO2 gate stack, such as lower bulk traps, interface

states, mobility and the resultant higher driving current, but also reduce the gate leakage

current substantially.

Experimental

The pMOSFETs were fabricated on n-type (100) 150 mm wafers. After conventional

LOCOS isolation, standard RCA cleaning with a final HF-dip was performed, followed by the

growth of an intentional 0.6nm thin interfacial oxyntride layer (SiON) using rapid thermal

processing in a NO2 ambient at 700°C. Subsequently, an approximately 3nm HfO2 layer was

substrate temperature of 500°C. The physical thicknesses of the SiON layer and HfO2 film

were measured by the optical n&k analyzer. After deposition of the HfO2 films, some samples

were subjected to an additional N2O-gas, NH3-gas or N2-gas plasma treatment at the substrate

temperature of 300°C. Then, all samples were annealed in a N2 ambient at 600°C for 30s in

order to improve the film quality. A 250nm polycrystalline silicon (poly-Si) layer was directly

deposited by low pressure chemical vapor deposition (LPCVD) on top of the HfO2 films, and

then the gate electrode patterning was implemented through lithography and etching

processing. Subsequently, the extension and deep source/drains were formed by implantation,

which dopants were activated at 950°C with rapid thermal annealing (RTA) for 20s in a N2

atmosphere. After passivation, contact holes formation, Al metallization and patterning, the

forming gas annealing at 400°C was finally performed for 30minutes.

Fig. 2 and 3 shows the gate leakage current of PMOSFET with HfO2 gate stack

subjected to various post-deposited plasma treatment under inversion and accumulation

region, respectively. The gate leakage current could be reduced by N2O plasma treatment both

at inversion and accumulation region, especially at inversion region, reduction can be

achieved two order magnitude.

Fig. 4 shows the typical C-V characteristics of HfO2 gate stack with various plasmas

lower than that at strong accumulation region, which is due to the poly depletion effect. In

addition, the hump occurs at the depletion region for all samples. This may be due to the

interface states and it seems the N2O-treated sample exhibits no significant hump. Fig. 5

shows the EOT extracted from strong inversion for HfO2 gate stack with various plasma

treatment. The EOT for N2O-treated sample increases, but for NH3-treated and N2-treated

samples decrease. The slightly EOT for N2O-treated sample may be one of the reason for

reduced leakage current.

The values of subthreshold swing of pMOSFETs with HfO2 gate stack objected various

plasma treatment. It can be found that N2O plasma-treated sample shows smaller swing

resulted from lower interface states. However, the degradation of swing are observed for

NH3-treated and N2-treated samples. This may be due to the ion bombardment from the

plasma.

The influences of various plasma treatments on the driving current of the pMOSFETs

with HfO2 gate stack. The driving current can be enhanced by post-deposited N2O plasma

treatment for HfO2 gate stack with higher EOT. Those results may be ascribed to lower

interface states and higher normalized transconductance, as shown in Fig.7. However, the

higher driving current for NH3-treated and N2-treated samples is attributed to smaller EOT

In order to gain into investigations of the improvements for N2O-treated samples, the

charge pumping measurements are used to analyze the interface states and bulk traps in the

dielectrics. Recently, the CP measurement had been frequently employed to qualify the level

of bulk traps 12,13) in the HfO2 dielectrics using the fixed base sweep and/or the fixed peak

sweep, as indicated in Fig. 8(b) and Fig. 8(c), respectively. The measured results of charge

pumping measurements are shown in the Figure 9. We can find that the interface states and

bulk traps 12,13) in the HfO2 dielectrics can be reduced by the post-deposition N2O plasma

treatment, resulted in the higher driving current and lower interface states even that it has

higher EOT.

Reference

1) E. A. Kerber et al., 2003 IRPS, pp.41.

2) S. Zafar, A. Callegari, E. gusev, and M. Fischetti: J. Appl. Phys. 93 (2003) 9298.

3) E. Gusev, D. A. Buchanan, E. Cartier: Int. Electron Device Meet. Tech. Dig., 2001, p. 451.

4) H. -J. Cho, C. Y. Kang, C. S. Kang, R. Choi, Y. H. Kim, M. S. Akbar, C. H. Choi, S. J.

Rhee, and J. C. Lee: IEEE Semiconductor Device Research Symposium, 2003, p. 68.

5) Y. Morisaki, T. Aoyama, Y. Sugita, K. Irino, T. Sugii, and T. Nakamura: Int. Electron

Devices Meet. Tech. Dig., 2002, p. 861.

Nishiyama: Int. Electron Device Meet. Tech. Dig., 2001, p. 459.

7) M. Koyama, A. Kaneko, T. Ino, M. Koike, Y. Kamata, R. Iijima, Y. Kamimuta, Takashima,

M. Suzuki, C. Hongo, S. Inumiya, M. Takayanagi and A. Nishiyama: Int. Electron

Devices Meet. Tech. Dig., 2002, p. 849.

8) R. Choi, C. S. Kang, B. H. Lee, K. Onishi, R. Nieh, S. Gopalan, E. Dharmarajan, and J. C.

Lee: Symp. VLSI Tech. Dig., 2001, p. 15.

9) M. Koyama, A. Kaneko, T. Ino, M. Koike, Y. Kamata, R. Iijima, Y. Kamimuta, Takashima,

M. Suzuki, C. Hongo, S. Inumiya, M. Takayanagi and A. Nishiyama: Int. Electron

Devices Meet. Tech. Dig., 2002, p. 849.

10) K. Onishi, S. K. Chang, R. Choi, Hag-Ju Cho; Gopalan, S, R. E. Nieh, S. A. Krishnan, and

J. C. Lee: IEEE Trans. Electron Devices 50 (2003) 384.

11) H. -J. Cho, C. S. Kang, K. Onishi, S. Gopalan, R. Nieh, R. Choi, E. Dharmarajan, J. C.

Lee: Int. Electron Device Meet. Tech. Dig., 2001, p. 655.

12) A. Kerber, E. Carter, L. Pantisano, M. Rosmeulen, R. Degraeve, T. Kauerauf, G.

Groeseneken, H. E. Maes, and U. Schwalke: Proc. 41th Int. Reliability Physics Symp,

Texas, 2003, p. 41.

13) A. Kerber, E. Carter, L. Pantisano, M. Rosmeulen, R. Degraeve, G. Groeseneken, H. E.

Fig.1 Process flow of HfO2 pMOSFET。

Fig. 2 The gate leakage at inversion region for various gas plasma treatments.

V_g (V) -6 -5 -4 -3 -2 -1 0 Ig (A) 10-13 10-12 10-11 10-10 10-9 10-8 10-7 10-6 As-dept N2O NH3 N2 PMOSFET L/W=2/10µm

V

(V)

0

1

2

3

4

5

6

I

(A

)

10

10

10

10

10

10

10

10

10

Fig. 3 The gate leakage at accumulation region for various gas plasma treatments.

Gate Voltage (V)

-3

-2

-1

0

1

2

Ca

p

a

ci

ta

n

c

e

(p

F)

3

4

5

6

7

8

9

Fig. 4 CV characteristics of HfO2 pMOSFET for various plasma treatments.

Split 0 1 2 3 4 5 6 EOT ( o A) 17 18 19 20 21 22 As-dept N2O NH3 N2

Fig. 5 The EOT values of HfO2 gate stack with various plasma treatment.

Method As- dept N2O NH3 N2 Swing value ~81 ~76 ~87 ~87

Table. 1 The swing values of HfO2 gate stack with various plasma treatment.

Fig. 8 The methods of various charge pumping measurements.(a) Fixed amplitude (b) Fixed Base (c) Fixed peak。

(a)

(b)

(c)

Fig. 9(c) Charge pumping current measured by Fixed base Method.

Gate Voltage (V) -2 -1 0 1 2 3 Icp ( A ) 0.0 5.0e-8 1.0e-7 1.5e-7 2.0e-7 2.5e-7 3.0e-7 As-dept N₂O NH₃ N₂ 4x10 µm Gate Voltage (V) -1.5 -1.0 -0.5 0.0 I d (1 0 -4 A) 0.5 1.0 1.5 2.0 2.5 3.0 _As-dept N2O NH3 N2

Fig. 6 The output current of HfO2 gate stack with various plasma treatment.

(V_g-V_Th)/EOT (MV/cm) -6 -4 -2 0 2 4 Gm x EOT (10 -4 Sxnm) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 As-dept N2O NH3 N2

Fig. 7 The normalized Gm of HfO2 gate stack with various plasma treatment

Gate Voltage (V) -3 -2 -1 0 1 2 3 Icp ( n A ) 0 25 50 75 100 125 As-dept N₂O NH₃ N₂ 4x10 µm

Fig. 9(a) Charge pumping current measured by Fixed amplitude Method.

Gate Voltage (V) -4 -2 0 2 Icp ( n A) 0 50 100 150 200 As-dept N₂O NH₃ N₂ 4x10 µm

Fig. 9(b) Charge pumping current measured by Fixed peak Method.